Suction imitation positive and negative pressure breast pump

The invention relates to a suction imitation positive and negative pressure breast pump, which can effectively improve milk discharge effect and solve the problem of milk accumulation. In the breast pump, the side face of a shell of a negative pressure cabin is provided with a milk suction disk port; the port is provided with a milk suction disk; an air bag is arranged in the milk suction disk and communicated with an exhaust hole; the lower part of the shell of the negative pressure cabin is provided with a milk bottle; a positive pressure cabin elastic piston is arranged in the negative pressure cabin; an upper opening part of the negative pressure cabin and an upper opening part of a positive pressure cabin are provided with cabin covers; a lifting rod which is extended out of the upper part of the cabin cover is arranged at the inner bottom of the positive pressure cabin; the negative pressure cabin, the positive pressure cabin, the lifting rod and the cabin covers form a positive and ...

Railway sand-proof and dust-proof monitoring system

The invention discloses a railway sand-proof and dust-proof monitoring system in the field of railways. The railway sand-proof and dust-proof monitoring system comprises an environment monitoring assembly, a sand burying depth monitoring assembly and a data processing module. According to information obtained by the environment monitoring assembly, after the information is processed by the data processing module, railway section information possibly affected by sand and dust is obtained, the corresponding number of sand-proof and dust-proof assemblies are installed on the corresponding sections, and the sand-proof and dust-proof assemblies are arranged in advance so that sand and dust hazards of railways can be prevented; and the damage of sand and dust to the driving safety of the railway can be avoided to a certain extent. The sand-proof and dust-proof assembly is used for obtaining the sand burial depth of the railway and transmitting the information of the sand burial depth of the railway ...

APPLYING LEARNING-TO-RANK FOR SEARCH

Techniques for applying learning-to-rank with deep learning models for search are disclosed herein. In some embodiments, a computer system trains a ranking model using training data and a loss function, with the ranking model comprising a deep learning model and being configured to generate similarity scores based on a determined level of similarity between profile data of reference candidates users in the training data and reference query data of reference queries in the training data. The computer system receives a target query comprising target query data from a computing device of a target querying user, and then generates a corresponding score for target candidate users based on a determined level of similarity between profile data of the target candidate users and the target query data using the trained ranking model. 1. A computer-implemented method comprising:receiving, by a computer system having a memory and at least one hardware processor, training data comprising a plurality of reference queries, a plurality of reference search results for each one of the plurality of reference queries, a plurality of user actions for each one of the plurality of reference queries, and a corresponding reaction indication for each one of the plurality of user actions, each one of the plurality of reference queries comprising reference query data and having been submitted by a reference querying user, the corresponding plurality of reference search results for each one of the plurality of reference queries comprising a plurality of reference candidate users displayed in response to the plurality of reference search queries based on profile data of the plurality of reference candidate users stored on a database of an online service, the plurality of user actions comprising actions by the reference querying user directed towards at least one reference candidate user of the plurality of reference search results for the corresponding reference search query, each reaction ...

DYNAMIC CANDIDATE POOL RETRIEVAL AND RANKING

Disclosed in some examples are methods, systems, and machine readable mediums which provide for retrieval, ranking, and display of candidates that are more likely to respond to employment inquiries in an employment search graphical user interface (GUI). The system may employ a machine learning algorithm which may calculate a score for each member of the social networking service that predicts, based upon one or more features how likely the individual is to respond to a message. In some examples, the candidates that are determined to be more likely to respond may be presented as a selectable option in the GUI. 1. A method for providing a graphical user interface (GUI) , the method comprising: ["for each particular member of a plurality of members of the social networking service, determining, using a machine learning algorithm, a response score for the particular member based upon one or more of: activity data describing the particular member's activity on the social networking service, and profile data, the response score indicative of a likelihood of the particular member to respond to an inquiry about a job from another member of the social networking service;", 'causing a search GUI to be displayed to a user, the search GUI including GUI elements that receive search criteria;', 'receiving through the search GUI, a search query with member search criteria;', 'responsive to the search query, identifying a set of members from the plurality of members based on the response score determined for each particular member from the set of members; and', 'causing the set of members to be displayed to the user as search results for the search query., 'using one or more processors of a social networking service, executing operations of2. The method of claim 1 , wherein receiving through the search GUI claim 1 , the search query with member search criteria comprises:receiving the member search criteria;determining a set of relevant members based upon the member search criteria; ...

Method for detecting illegal tower climbing behavior of electric power maintainer based on deep learning

The invention provides a method for detecting illegal tower climbing behaviors of electric power maintainers based on deep learning. The method is used for solving the technical problems of requirements of tower climbing behavior detection of electric power maintainers based on deep learning on measurement environment and equipment and improvement of detection accuracy in the prior art. The method comprises the following steps: (1) collecting limb posture data of an electric power maintainer; (2) constructing an illegal tower climbing behavior detection MKeras deep neural network formed by connecting a feature extraction module and a classification module in series; (3) training an MKeras deep neural network; and (4) judging the tower climbing behavior of the electric power maintenance personnel. According to the method, the MKeras deep neural network is established, so that the detection of the illegal tower climbing behavior of the electric power maintainer by the MKeras deep neural network ...

Efficient simulation method and device for power system, equipment and storage medium

The embodiment of the invention discloses a power system efficient simulation method and device, equipment and a storage medium, the method is applied to a real-time simulation system, the system comprises a first power system and a second power system, and the method comprises the steps that the first power system is controlled to conduct electromagnetic transient simulation according to a first step length, and the current simulation frequency is determined; and obtaining a target element corresponding to the current simulation frequency according to the corresponding relationship between the current simulation frequency and each element in the second power system, generating a simulation instruction, sending the simulation instruction to the second power system, and indicating the second power system to execute electromechanical transient simulation on the target element. By counting the operation times of the first power system, the target element corresponding to the current simulation ...

Apparatus and method for performing a forward operation of artificil neural networks

Aspects for executing forward propagation of artificial neural network are described here. As an example, the aspects may include a plurality of computation modules connected via an interconnection unit; and a controller unit configured to decode an instruction into one or more groups of micro-instructions, wherein the plurality of computation modules are configured to perform respective groups of the micro-instructions.

Single local oscillator in a multi-band frequency division duplex transceiver

Embodiments of the present disclosure relate to multi-band FDD transceivers. An example transceiver includes a LO, configured to generate a LO signal to be shared between a receiver and a transmitter of the transceiver. Both the receiver and the transmitter use quadrature signal processing and are configured to multi-band operation. Sharing a single LO to perform frequency conversion of different frequency bands of received and transmitted signals advantageously allows reducing the number of LOs used in a multi-band FDD transceiver.

STABILIZED PHOTO-ALIGNMENT LAYER FOR LIQUID CRYSTAL

A liquid-crystal photo-alignment layer that is composed of polymer stabilized azo dyes is provided, where a polymer network is introduced in the photo-alignment layer for stabilization. The photo-alignment layer is realized based on a two-step irradiation to first achieve molecule alignment to form the photo-alignment layer and then stabilize this layer. To realize the photo-alignment layer, a pre-determined surface of the substrate is first coated with a film of mixture. The mixture comprises an azo dye and a monomer preferably mixed in an optimal concentration of 0.67 wt/wt. The azo dye and the monomer have light-absorption peaks at different wavelengths such that photo-alignment of the azo-dye molecules and stabilization of the photo-alignment layer by polymerization of the monomer are achievable by two separate exposures of light to the film.

HVAC HYBRID BLOWER MOTOR SOFT START

A method for soft start of a motor in a heating, ventilation, and cooling (HVAC) system. The method includes operably connecting a first switching device with the motor; the first switching device operably connected to a power source and configured to direct power from a DC power source to the motor, and operably connecting a second switching device with the motor, the second switching device operably connected to a power source and configured to direct power from the DC power source to the motor. The method also includes receiving, by a controller, a request to engage the motor associated with an operation of the HVAC system, commanding the first switching device with a first pulse width modulation signal for a first selected duration, while ensuring the second switching device is disengaged, and after the first selected duration, enabling the second switching device and disengaging the first switching device. 1. A method for soft start of a blower motor in a heating , ventilation , and cooling (HVAC) system , the method comprising:operably connecting a first switching device with the motor; the first switching device operably connected to a DC power source and configured to direct power from the DC power source to the motor;operably connecting a second switching device with the motor, the second switching device operably connected to the DC power source and configured to direct power from the DC power source to the motor;receiving a request to engage the motor associated with an operation of the HVAC system;commanding the first switching device with a first pulse width modulation (PWM) signal for a first selected duration, while ensuring the second switching device is disengaged, the commanding the first switching device with the first PWM signal comprising applying the first PWM signal to the first switching device causing the first switching device to activate at a selected duty cycle of the first PWM signal; andafter the first selected duration, enabling the ...

Apparatus and Methods for Vector Operations

Aspects for vector operations in neural network are described herein. The aspects may include a vector caching unit configured to store a vector, wherein the vector includes one or more elements. The aspects may further include a computation module that includes one or more comparers configured to compare the one or more elements to generate an output result that satisfies a predetermined condition included in an instruction. 1. An apparatus for vector operations in a neural network , comprising:a controller unit configured to receive a vector operation instruction that indicates an address of a vector and a predetermined condition; and wherein the vector includes one or more elements, and', 'wherein the computation module includes one or more comparers configured to compare the one or more elements to generate an output result that satisfies the predetermined condition included in the vector operation instruction., 'a computation module configured to receive the vector based on the address included in the vector operation instruction,'}2. The apparatus of claim 1 ,wherein the vector operation instruction further indicates a length of the vector, andwherein the computation module is configured to retrieve the vector based on the length of the vector and the address of the vector.3. The apparatus of claim 1 , wherein the vector operation instruction includes one or more register IDs that identify one or more registers configured to store the address of the vector and a length of the vector.4. The apparatus of claim 1 , wherein the comparers are configured to select a maximum element from the one or more elements as the output result.5. The apparatus of claim 1 , wherein the comparers are configured to select a minimum element from the one or more elements as the output result.6. The apparatus of claim 1 , wherein the controller unit comprises an instruction obtaining module configured to obtain the vector operation instruction from an instruction storage device.7. ...

Optimization method and apparatus for verification vectors

The invention discloses an optimization method and a device for verifying a vector. The method comprises the following steps: according to the simulated initial verification vector, obtaining information on coverage rate of function points covered by the initial verification vector, initializing a list of the covered function points, and constructing a classification model; and sending a generated verification vector to the classification model, predicting the coverage rate of the function points covered by the generated verification vector, simulating the generated verification vector for improving the coverage rate of the function points, and updating the list of the function points according to the simulation result to obtain the verification vector with higher coverage rate. The method can reduce the number of the verification vector needing to be simulated, reduce the usage of the verification vector as much as possible and shorten simulation time, so as to improve verification efficiency ...

Method and device for discovering interest of users

The invention discloses a method and device for discovering interest of users. The method mainly includes the steps that network access behavior data of users are acquired; a field to which the network access behavior data belong is determined according to entity words contained in the network access behavior data and multiple preset entity words corresponding to fields respectively; weighted values of the network access behavior data are calculated according to multi-dimensional attribute information corresponding to the network access behavior data; the attention of the users to the field to which the network access behavior data belong is determined according to the weighted values of the network access behavior data of the users; the interest of the users is recognized according to the attention to the field to which the network access behavior data belong and a preset interest threshold value corresponding to the corresponding field, wherein the interest threshold value corresponding ...

System and method for three-dimensional characterization of blade tip gaps and dynamic detection of optical fibers of aero-engine

The invention discloses a system and a method for three-dimensional characterization of blade tip gaps and dynamic detection of optical fibers of an aero-engine. The system comprises a laser light source, a pulsed light source control circuit, a light splitter, a blade tip gap sensing probe, a pre-processing circuit, a single chip and an upper computer; the blade tip gap sensing probe comprises a reference optical fiber, an axial angle-measuring optical fiber and a circumferential angle-measuring optical fiber. The system is capable of dynamically and accurately obtaining the three-dimensional characterization of blade tip gaps of the aero-engine and is high in anti-interference capability.

Optical and SAR image automatic registration method based on bidirectional style migration

The invention discloses an optical and SAR image automatic registration method based on bidirectional style migration, and belongs to the technical field of remote sensing image intelligent processing. According to the method, firstly, optical and SAR images are converted into pseudo-SAR and pseudo-optical images through a CycleGAN style migration network, homogenization of heterogenous images is achieved, then feature point sets are extracted from the optical and SAR images and the converted pseudo-SAR and pseudo-optical images respectively through SIFT, a feature point set of an optical image to be registered and a feature point set of an SAR image to be registered are obtained, and the feature point set of the SAR image to be registered is obtained. Extracting an optical mixed descriptor and an SAR mixed descriptor, calculating the approximation degree of the mixed descriptors, screening a feature point set, and finally obtaining matched feature point pairs through RANSAC outer point ...

Method for testing penetrating power of industrial CT (Computed Tomography) system

The invention relates to an industrial CT (computed tomography) system penetrability testing method which comprises the following steps: designing a test block for testing a tested industrial CT system, the test block is a cylinder and is internally provided with a through hole coaxially arranged with the cylinder, and the longitudinal section of the through hole is trapezoidal; placing the test block at the central position of a rotary table, and performing CT scanning imaging on the test block by using a tested industrial CT system to obtain a plurality of test block CT images; obtaining the material center position of each test block CT image and the penetration thickness of each test block CT image; calculating a point spread function of the CT image; obtaining a gray value of an internal gap part and a gray value of a material part in each test block CT image; calculating contrast values corresponding to different spatial resolutions under different penetration thicknesses; and when ...

Polymer Stabilized Electrically Suppressed Helix Ferroelectric Liquid Crystal Cell

The present invention provides an electrically suppressed helix ferroelectric liquid crystal (ESHFLC) cell with polymer stabilization. The cell has a liquid crystal (LC) material that is a mixture comprising a monomer, a photo-initiator and a ferroelectric liquid crystal (FLC) where a polymer network has been established at a certain temperature to achieve constraints of ESHFLC electro-optical mode. The resultant mixture is characterized by a helix pitch less than and comparable to a FLC layer thickness of the cell, and provides a selective reflection in an UV region. The concentration of the monomer in the pure FLC mixture has also been optimized for the phase diagram, scattering and the tilt angle. The resultant mixture, i.e. the polymer stabilized ESHFLC cell, follows all constraints of the ESHFLC electro-optical mode and shows electro-optical characteristics similar to a typical ESHFLC cell. 1. A polymer stabilized electrically suppressed helix ferroelectric liquid crystal cell comprising a composite ferroelectric liquid crystal formed by a blended mixture of a monomer , a photo initiator and a pure ferroelectric liquid crystal , the monomer being polymerized such that a polymer network is formed in the cell , wherein:the pure ferroelectric liquid crystal is stabilized by the polymer network in order that the composite ferroelectric liquid crystal has a helix having a pitch less than and comparable to a ferroelectric liquid crystal layer thickness of the cell as well as provides selective reflection in an ultraviolet (UV) region, whereby an elastic energy of the helix is comparable to an anchoring energy of aligning substrates of the cell.2. The cell of claim 1 , wherein the cell is prepared by mixing the monomer in the pure ferroelectric liquid crystal in an optimal concentration of having less than 10% of monomer in the pure ferroelectric liquid crystal.3. The cell of claim 1 , wherein the monomer is polymerized after heating the blended mixture at an optimum ...

Gossypium anomalum-sourced ssr sequence associated with high lint percentage and drought tolerance and application thereof

Provided is a gossypium anomalum-sourced SSR sequence associated with high lint percentage and drought tolerance and an application thereof. Provided is a gossypium anomalum-sourced chromosome segment which is a DNA segment from an SSR molecular marker JAAS6365 to an SSR molecular marker JAAS5604 on chromosome 5 of the gossypium anomalum. The nucleotide sequence of the SSR molecular marker JAAS6365 is shown as SEQ ID No. 1, and the nucleotide sequence of the SSR molecular marker JAAS5604 is shown as SEQ ID No. 2.

SCHEDULING IN JOB EXECUTION

The present invention relates to a method, apparatus, and computer program product for scheduling in job execution. According to embodiments of the present invention, there is provided a method for scheduling a plurality of job slots shared by one or more pre-processors and one or more post-processors in job execution, wherein the data generated by the pre-processor(s) will be fed to the post-processor(s) for processing. The method comprises: determining an overall data generation speed of the pre-processor(s); determining an overall data consumption speed of the post-processor(s); and scheduling allocation of at least one of the job slots between the pre-processor(s) and the post-processor(s) based on the overall data generation speed and the overall data consumption speed. Corresponding apparatus is disclosed as well. 1. A method for scheduling in job execution , the method comprising:determining a data feed speed of a first post-processor, wherein the data feed speed of the first post-processor is a speed of feeding data from a plurality of pre-processors to the first post-processor;determining a data feed speed of a second post-processor, wherein the data feed speed of the second post-processor is a speed of feeding data from the plurality of pre-processors to the second post-processor;determining a local data feed speed of the first post-processor, wherein the local feed speed of the first post-processor is a speed of feeding data, to the first post-processor, from one or more pre-processors of the plurality of pre-processors that are co-located with the first post-processor on a physical node of the first post-processor;determining a local data feed speed of the second post-processor, wherein the local feed speed of the second post-processor is a speed of feeding data, to the second-processor, from one or more pre-processors of the plurality of pre-processors that are co-located with the second post-processor on a physical node of the second post-processor; ...

APPARATUS AND METHOD FOR EXECUTING RECURRENT NEURAL NETWORK AND LSTM COMPUTATIONS

Aspects for Long Short-Term Memory (LSTM) blocks in a recurrent neural network (RNN) are described herein. As an example, the aspects may include one or more slave computation modules, an interconnection unit, and a master computation module collectively configured to calculate an activated input gate value, an activated forget gate value, a current cell status of the current computation period, an activated output gate value, and a forward pass result. 1. An apparatus for backward pass in a recurrent neural network (RNN) network , comprising:one or more slave computation modules configured to calculate a first cell output partial sum and a second cell output partial sum; andan interconnection unit configured to add the first cell output partial sum and the second cell output partial sum to generate one or more cell output gradients.2. The apparatus of claim 1 , further comprising a master computation module configured to activate a dormant output gate value with a derivative of an activation function to generate an activated output gate value claim 1 ,wherein the one or more slave computation modules are configured to multiply the cell output gradients with activated current cell status to generate a cell output multiplication result, andwherein the master computation module is further configured to multiply the activated output gate value with the cell output multiplication result to generate one or more output gate gradients.3. The apparatus of claim 2 ,wherein the one or more slave computation modules are configured to calculate a first cell status partial sum, a second cell status partial sum, a third cell status partial sum, a fourth cell status partial sum, and a fifth cell status partial sum; andwherein the interconnection unit is configured to add the first cell status partial sum, the second cell status partial sum, the third cell status partial sum, the fourth cell status partial sum, and the fifth cell status partial sum to generate one or more cell ...

Scheduling in job execution

The present invention relates to a method, apparatus, and computer program product for scheduling in job execution. According to embodiments of the present invention, there is provided a method for scheduling a plurality of job slots shared by one or more pre-processors and one or more post-processors in job execution, wherein the data generated by the pre-processor(s) will be fed to the post-processor(s) for processing. The method comprises: determining an overall data generation speed of the pre-processor(s); determining an overall data consumption speed of the post-processor(s); and scheduling allocation of at least one of the job slots between the pre-processor(s) and the post-processor(s) based on the overall data generation speed and the overall data consumption speed. Corresponding apparatus is disclosed as well.

Mobile terminal intelligent information analysis and auxiliary decision-making system

The invention relates to a mobile terminal intelligent information analysis and auxiliary decision-making system, which comprises a threat evaluation module, a path planning module and a trajectory prediction module, and is characterized in that the threat evaluation module is used for threat value evaluation of a combat unit target and threat value evaluation of a specified combat area, and visual display of threats is realized; the path planning module is used for planning and displaying a path and realizing a planning task leading to a target position; and the trajectory prediction module is used for situation visualization display, screening of situation analysis targets, trajectory prediction parameter setting, target future position deduction display, target action trajectory prediction and deduction display of target future moments. According to the method, front-line personnel in field tasks can conveniently and visually complete conventional intelligent information analysis and ...

Smart suggestions personalization with GLMix

In an example, an indication of a plurality of different entities in a social networking service is received, including at least two entities having a different entity type. Then a plurality of user profiles in the social networking service are accessed. A machine-learned model is then used to calculate, based on co-occurrence counts reflecting a number of user profiles in the plurality of user profiles in which corresponding nodes co-occurred, a similarity score between a first node and second node by computing distance between the first node and the second node in a d-dimensional space on which a plurality of entities are mapped, the similarity score generated using a generalized linear mixed model having a global coefficient vector applied to global function pertaining to the co-occurrence counts and a first random effects coefficient vector applied to a random effects per-country function.

Information processing method and electronic device

The invention provides an information processing method and an electronic device. The method is applied to a solid-state storage device which is connected with the electronic device; the solid-state storage device is capable of supporting N different logical address-physical address mapping tables at the same time, wherein N is an integer which is more than or equal to 1; the method comprises the steps of receiving identity information of a user from the electronic device; determining a first logical address-physical address mapping table corresponding to the user based on the identity information, and assigning the first logical address-physical address mapping table to the user.

Active power distribution network-oriented reactive voltage support and fault active regulation and control composite device

The reactive voltage support and fault active regulation and control composite device for the active power distribution network comprises an in-station capacitor unit, an active inversion unit and an in-station arc suppression coil unit, the in-station capacitor unit and the active inversion unit are respectively hung on the power distribution network, and the output end of the active inversion unit is grounded through the in-station arc suppression coil unit. When a power grid operates normally, the device works in a reactive compensation mode, and the in-station capacitor unit and the active inverter unit provide reactive power needing to be compensated for a load; when a single-phase earth fault occurs in the power grid, the device works in a fault regulation and control mode, and the single-phase inverter and the arc suppression coil of a non-fault phase cooperate with each other to inject arc suppression current into the power distribution network together so as to suppress the fault ...

MULTIPLICATION AND ADDITION DEVICE FOR MATRICES, NEURAL NETWORK COMPUTING DEVICE, AND METHOD

Aspects of a neural network operation device are described herein. The aspects may include a matrix element storage module configured to receive a first matrix that includes one or more first values, each of the first values being represented in a sequence that includes one or more bits. The matrix element storage module may be further configured to respectively store the one or more bits in one or more storage spaces in accordance with positions of the bits in the sequence. The aspects may further include a numeric operation module configured to calculate an intermediate result for each storage space based on one or more second values in a second matrix and an accumulation module configured to sum the intermediate results to generate an output value. 1. A neural network operation device , comprising:a submatrix divider circuit configured to select a portion of an input data matrix as an input submatrix; receive a convolution kernel matrix that includes one or more kernel values, wherein each of the one or more kernel values is represented as a sequence that includes one or more bits, and', 'respectively store the one or more bits in one or more storage spaces in accordance with positions of the one or more bits in the sequence;, 'a matrix element memory configured toa calculator circuit configured to calculate an intermediate result for each storage space based on one or more input elements in the input submatrix, wherein the one or more input elements correspond to non-zero values stored in the storage space;an accumulator circuit configured to sum the intermediate results to generate an output value; anda convolution result assembler circuit configured to assemble the output values calculated for different portions of the input data matrix to generate an output matrix.2. The neural network operation device of claim 1 , wherein the submatrix divider circuit is further configured to select the input submatrix in accordance with the convolution kernel matrix.3. The ...

CONVERSION METHOD, APPARATUS, COMPUTER DEVICE, AND STORAGE MEDIUM

A model conversion method, apparatus, computer device, and storage device are disclosed. The method can convert an initial offline model to a target offline model. The model conversion method, apparatus, computer device, and storage device may greatly reduce data input of the computer device and have a simple conversion process, and may thus reduce an amount of data processing of the computer equipment, improve processing efficiency and reduce power consumption.

Method for post extraction of low-acyl gellan gum with high solubility

The invention relates to a method for post extraction of low-acyl gellan gum with high solubility, which comprises the following steps: (1) enzyme treatment of fermentation liquor; (2) flocculation of the fermentation liquor; (3) acyl removal treatment; (4) clarification of the solution after the acyl removal; (5) dehydration of a clarified acyl-removed gellan gum solution; (6) ion exchange and decoloring treatment; and (7) drying and pulverization. The product obtained by the method has the advantages of greatly-improved solubility, greatly-improved quality, good appearance, high transmittance and high gel strength.

Nano material based on molybdenum disulfide network structure as well as preparation method and application of nano material

The invention provides a nanometer material based on a molybdenum disulfide network structure and a preparation method thereof.The method comprises the steps that an aqueous solution of cysteine and an aqueous solution of anhydrous sodium molybdate are mixed and then subjected to ultrasonic treatment, and a molybdenum disulfide precursor solution is obtained; the molybdenum disulfide precursor solution is subjected to a hydrothermal reaction, after the reaction is finished, cleaning and centrifugation are conducted, and a molybdenum disulfide nanosphere material is obtained; and adding the molybdenum disulfide nanosphere material into a toluene solution, carrying out ultrasonic treatment, and then carrying out centrifugal cleaning to obtain the nanomaterial with the molybdenum disulfide nanosphere network structure. The nanometer material based on the molybdenum disulfide network structure is used for removing formaldehyde. Molybdenum disulfide is adopted, the molybdenum disulfide is a ...

Spare wheel carrier of trailer

The trailer spare tire carrier comprises a supporting mounting frame, a supporting seat is arranged at one end of the supporting mounting frame, and an arc-shaped seat is arranged at the other end of the supporting seat; a first supporting plate is rotationally connected to one side of the supporting seat, a second supporting plate is rotationally connected to the end, away from the supporting seat, of the first supporting plate, and a third supporting plate is rotationally connected to the end, away from the first supporting plate, of the second supporting plate; a clamping base is arranged at the end, away from the second supporting plate, of the third supporting plate. The arc-shaped base, the supporting base, the first supporting plate, the second supporting plate, the third supporting plate and the clamping base surround the outer portion of a placed spare tire. The spare tire can be protected and damped conveniently, and meanwhile the spare tire can be taken down and placed conveniently ...

DEVICE FOR SAMPLE COLLECTION, TRANSPORTATION, AND PROCESSING

A device system, and method of use thereof, for collecting, transporting, and processing a biological sample is provided. 1. A device system for collecting , transporting , and processing a biological sample , comprising:a) a sample collection assembly;b) a collection extraction basket,c) a place holder, andd) a hinged holder.2. The device system of claim 1 , wherein said sample collection assembly comprises:a) a swab,b) a cap, andc) a handle comprising a sample identifier.3. The device system of claim 2 , wherein said swab claim 2 , cap claim 2 , and handle claim 2 , each can be detachable from each other.4. The device system of claim 2 , wherein said sample identifier is a barcode.5. The device system of claim 1 , wherein said collection extraction basket also comprises a barcode or other readable identifier which matches the sample identifier labeled on either the cap or the handle of the detachable sample collection assembly.6. The device system of claim 2 , wherein said collection extraction basket comprises multiple tubes having small holes at the bottom that allow for solution to enter the collection extraction basket.7. The device system of claim 1 , wherein said place holder system is made of polypropylene or other suitable polymers claim 1 , glass claim 1 , and/or plastic materials.8. The device system of claim 7 , wherein said place holder is a plastic cover comprising wells corresponding a microtiter plate with the same number of well spaces.9. The device system of claim 1 , wherein said hinged holder comprising multiple number of spaces corresponding to a microtiter plate with the same number of well spaces.10. The device of claim 9 , wherein said hinged holder comprises two positions to lower or raise said sample collection assembly with the collection extraction basket.11. The device system of claim 1 , wherein said biological sample comprises tissue claim 1 , blood claim 1 , urine claim 1 , faces claim 1 , skin claim 1 , hair claim 1 , or any other ...

METHOD AND APPARATUS FOR TASK SCHEDULING ON HETEROGENEOUS MULTI-CORE RECONFIGURABLE COMPUTING PLATFORM

A method for task scheduling on a heterogeneous multi-core reconfigurable computing platform is disclosed, the method includes: determining an execution manner of a target task according to popularity of the to-be-executed target task and usage of a reconfigurable resource of the heterogeneous multi-core reconfigurable computing platform, where the execution manner includes a hardware manner or a software manner, for the target task, there is no corresponding target intellectual property IP core for executing the target task on the heterogeneous multi-core reconfigurable computing platform, the popularity of the target task is used to indicate central processing unit CPU usage of the target task, and the usage of the reconfigurable resource is used to indicate a usage status of the reconfigurable resource; and executing the target task according to the determined execution manner. 1. A method for task scheduling on a heterogeneous multi-core reconfigurable computing platform , the method comprising:determining an execution manner of a target task according to popularity of the target task which is to-be-executed and usage of a reconfigurable resource of the heterogeneous multi-core reconfigurable computing platform, wherein the execution manner comprises a hardware manner or a software manner, for the target task, there is no corresponding target intellectual property (IP) core for executing the target task on the heterogeneous multi-core reconfigurable computing platform, the popularity of the target task is used to indicate central processing unit (CPU) usage of the target task, and the usage of the reconfigurable resource is used to indicate a usage status of the reconfigurable resource; andexecuting the target task according to the determined execution manner.3. The method according to claim 2 , wherein the determining the execution manner of the target task according to the weighted popularity Cof the target task and the usage Eof the reconfigurable resource ...

Multiplication and addition device for matrices, neural network computing device, and method

Aspects of a neural network operation device are described herein. The aspects may include a matrix element storage module configured to receive a first matrix that includes one or more first values, each of the first values being represented in a sequence that includes one or more bits. The matrix element storage module may be further configured to respectively store the one or more bits in one or more storage spaces in accordance with positions of the bits in the sequence. The aspects may further include a numeric operation module configured to calculate an intermediate result for each storage space based on one or more second values in a second matrix and an accumulation module configured to sum the intermediate results to generate an output value.

Apparatus and Methods for Training in Fully Connected Layers of Convolutional Networks

Aspects for backpropagation in a fully connect layer of a convolutional neural network are described herein. The aspects may include a direct memory access unit configured to receive input data and one or more first data gradients from a storage device. The aspects may further include a master computation module configured to transmit the input data and the one or more first data gradients to one or more slave computation modules. The slave computation modules are respectively configured to multiply one of the one or more first data gradients with the input data to generate a default weight gradient vector. 1. An apparatus for backpropagation in a fully connected layer of a neural network , comprising:a controller unit configured to receive an instruction; andone or more computation modules that include a master computation module and one or more slave computation modules, receive input data and one or more first data gradients in response to the instruction, and', 'transmit the input data and the one or more first data gradients to one or more slave computation modules, and, 'wherein the master computation module configured to'}wherein the one or more slave computation modules are respectively configured to multiply one of the one or more first data gradients with the input data to generate a default weight gradient vector.2. The apparatus of claim 1 , wherein the master computation module is further configured to update one or more weight values based on the default weight gradient vector.3. The apparatus of claim 1 , wherein the master computation module is further configured tocalculate a scaled weight gradient vector based on the default weight gradient vector and a predetermined threshold value; andupdate one or more weight values based on the scaled weight gradient vector.4. The apparatus of claim 1 , wherein the master computation module is further configured to apply a derivative of an activation function to the one or more first data gradients to generate ...

Method for Post-Extracting Low Acyl Gellan Gum with High Solubility

The invention relates to a post-extraction process for preparing low acyl gellan gum having high solubility property, the process comprising the following steps: (1) enzyme treatment of fermentation broth; (2) flocculation of fermentation broth; (3) deacylation treatment; (4) clarification treatment of deacylated solution; (5) dehydration treatment of clarified and deacylated gellan gum solution; (6) ion exchange and decoloration treatment; and (7) drying and milling. The process according to the invention produces a product, which has largely improved solubility, largely improved quality, good appearance, high transparence, and high gel strength.

Data processing method, recommendation source information organization, information recommendation method and information recommendation device

The invention discloses a data processing method, a recommendation source information organization, an information recommendation method and an information recommendation device, wherein the data processing method comprises the following steps of: filtering collected behavior recoding data for network access by a plurality of users so as to obtain sample data with categories form the collected recording data; for the sample data with the categories, carrying out feature word extraction on various categories of sample data; and storing the corresponding relationship between the extracted feature words and the corresponding categories.

Station simulation method and device, nonvolatile storage medium and computer equipment

The invention discloses a station simulation method and device, a nonvolatile storage medium and computer equipment. The method comprises the following steps: according to respective power characteristic parameters of a plurality of power generation units in a new energy station, determining respective coordinate positions of the plurality of power generation units in a power generation unit characteristic coordinate system; first distances between the multiple power generation units and a first center position are determined, and the first center position is a coordinate position corresponding to a power generation unit selected from the multiple power generation units in a power generation unit characteristic coordinate system; optimizing the first center position according to the first distance to obtain a target center position; and taking the target power generation unit corresponding to the target center position as an equivalent power generation unit of the plurality of power generation ...

Hand-held mobile apparatus and method of quckly displaying online community friends' images applicable thereto

A hand-held mobile apparatus and method of quickly displaying online community friends' images applicable thereto are disclosed. The hand-held mobile apparatus includes a network communication module for linking to an online community server; a comparison module for comparing and determining whether data of a web image in the online community server matches data of a local image in a storage unit of the apparatus; a recognition module for determining an image type of the web image when the web image data is different from the local image data; a processing unit for downloading the web image of a first type in a smallest image size or the web image of a second type in an image size matching a screen resolution of the apparatus, and storing the web image in the storage unit as the local image; and a display unit for displaying the local image.

Neural network convolution computation method and device, and computer-readable storage medium

Aspects of a neural network convolution device are described herein. The aspects may include a matrix transformer and a matrix multiplication module. The matrix transformer may be configured to receive an input data matrix and a weight matrix, transform the input data matrix into a transformed input data matrix based on a first transformation matrix, and transform the weight matrix into a transformed weight matrix based on a second transformation matrix. The matrix multiplication module may be configured to multiply one or more input data elements in the transformed input data matrix with one or more weight elements in the transformed weight matrix to generate an intermediate output matrix. The matrix transformer may be further configured to transform the intermediate output matrix into an output matrix based on an inverse transformation matrix.

PROCESSING APPARATUS AND PROCESSING METHOD

The present disclosure provides a processing device and method. The device includes: an input/output module, a controller module, a computing module, and a storage module. The input/output module is configured to store and transmit input and output data; the controller module is configured to decode a computation instruction into a control signal to control other modules to perform operation; the computing module is configured to perform four arithmetic operation, logical operation, shift operation, and complement operation on data; and the storage module is configured to temporarily store instructions and data. The present disclosure can execute a composite scalar instruction accurately and efficiently.

Transcription factor RNA interference reagents and methods of use thereof

Composite additive improving self-contraction of concrete

Clustering method, searching method and system for interconnection network audio file

Wind electricity active power control distribution method and system

The invention relates to a wind electricity active power control distribution method and system. The method includes the steps: acquiring wind electricity active power total adjustment amount; reading basic information needed by calculating comprehensive scores of wind electricity plants when the update period of the comprehensive scores of the wind electricity plants is achieved, calculating a comprehensive score of each wind electricity plant, and generating or updating a power reduction wind electricity plant queue and a power rising wind electricity plant queue according to the comprehensive score of each wind electricity plant; sequentially performing power reduction distribution for the wind electricity plants in the power reduction queue or power rising distribution for the wind electricity plants in the power rising queue according to the wind electricity active power adjusting direction. Real-time running data of the wind electricity plants are combined, scoring and sequencing ...

Overlapped gravity block underwater sand consolidation structure

The invention provides an overlapped type gravity block underwater sand consolidation structure. The overlapped type gravity block underwater sand consolidation structure comprises overlapped type gravity blocks and connecting components. The overlapped gravity block is in a shape with a thick middle part and a thin edge, so that a lap joint structure is formed at the edge when the block bodies are arranged, and a flexible water filtering material which does not allow sand to pass through is arranged in a transverse gap of a lap joint part; a steering groove for mounting a connecting component is reserved in the overlapped gravity block; the connecting component is composed of a penetrating type fiber expanding structure and a flexible fiber woven rope. Overlapped gravity blocks are arranged within the range of sand body movement caused by water body movement, the adjacent gravity blocks are connected into a whole through a connecting component, and a block body clear space is formed between ...

Detection method and device for position signal abnormity of switch knife switch

The invention provides a method and device for detecting position signal abnormity of a switch disconnecting link of a direct-current power transmission system, and the method comprises the steps: determining a reference signal value according to the number of first signal collection values and the number of second signal collection values; determining a plurality of comparison results according to the reference signal value and the plurality of acquisition signal values; and determining an abnormal flag bit according to a comparison result. Judging the collection value corresponding to the maximum value in the number of the first signal collection values and the number of the second signal collection values as the actual value of the switch knife switch position signal, and comparing the judged actual value of the switch knife switch position signal with the switch knife switch position signal collection values collected by each set of control protection device one by one; therefore, the ...

Fault diagnosis method and fault diagnosis device for power system and electronic device

The invention provides a fault diagnosis method and device of a power system and an electronic device. The method comprises the steps of performing data preprocessing and feature extraction on a received original data set to obtain a target data set; a model structure and hyper-parameters of a preset fault diagnosis model are determined based on a target algorithm, the preset fault diagnosis model is trained based on the target data set, a target fault diagnosis model is obtained, and the target algorithm comprises one of a genetic algorithm, a Bayesian optimization algorithm and a fault diagnosis algorithm. The preset fault diagnosis model is used for performing fault diagnosis on the target power system; and inputting the power system data of the target power system received in real time into the target fault diagnosis model to obtain a fault diagnosis result of the target power system, thereby solving the problem of low accuracy and efficiency caused by performing fault diagnosis on ...

Method and apparatus for task scheduling on heterogeneous multi-core reconfigurable computing platform

A method for task scheduling on a heterogeneous multi-core reconfigurable computing platform is disclosed, the method includes: determining an execution manner of a target task according to popularity of the to-be-executed target task and usage of a reconfigurable resource of the heterogeneous multi-core reconfigurable computing platform, where the execution manner includes a hardware manner or a software manner, for the target task, there is no corresponding target intellectual property IP core for executing the target task on the heterogeneous multi-core reconfigurable computing platform, the popularity of the target task is used to indicate central processing unit CPU usage of the target task, and the usage of the reconfigurable resource is used to indicate a usage status of the reconfigurable resource; and executing the target task according to the determined execution manner.

Near space balloon system safety release method

The invention discloses a near space balloon system safety release method. The method comprises an unfolding stage, an inflation stage and a let-fly stage. By adopting the method, the windward area of a balloon is small in the inflation stage, an obvious wind-catching phenomenon is not caused, a set of hoists is adopted for multi-point constraint, the state of the balloon is controlled, thus the balloon is fixed safely and reliably, the wind resistance ability is high, thus the whole release process of a balloon system can be completely carried out in an external field, and a shelter place or a large-area ground is not needed; a descent control device is adopted, so that the liftoff speed of a nacelle and the relative speed between the nacelle and the balloon are lowered and the balloon system is lifted off safely; in the release process of the balloon, release can be performed only under the condition that rope belts are at a straightened state between the balloon and the nacelle, the ...

Process for extracting matrine form Chinese traditional medicine extract

The invention discloses a method for extracting matrine from a sophora flavescens ait, which comprises three steps as follows: matrine molecular imprinting polymer is prepared, matrine molecular imprinting polymer is used as solid-phase extraction medium, and a matrine molecular imprinting solid-phase extraction column is used for extracting and separating matrine. A novel continuous rodlike molecule imprinting polymer which is synthesized by taking matrine as template molecules has the specific capacity for recognizing matrine; obtained polymer is prepared into the solid-phase extraction column through the means of packing or direct synthesis in the column; the method can be used for extraction, separation and purification of matrine; the obtained product has high purity quotient, the operating procedure is simple, the solvent usage is reduced, the extraction is safer and environment-friendly, and the purposes of removing impurities as well as purifying and enriching matrine can be served ...

Bottled water labeller production line bottle comdenstion water rapid removal device

The utility model discloses a bottled water labeller production line bottle comdenstion water rapid removal device, including the water collector, the upper end fixedly connected with protection casing of water collector, the upper end fixedly connected with roof of protection casing, the first axle bed of lower surface one side fixedly connected with of roof, the inboard hub connection of first axle bed has the gangbar, the other end of gangbar was equipped with the cab apron, the other end of hub connection gangbar is passed through to the upper end of crossing the cab apron, cross the firstair knife of one side fixedly connected with of cab apron. This bottled water labeller production line bottle comdenstion water rapid removal device, the body after the outlet pipe is accomplished the filling carries out the trickle, becomes water bead -shaped attitude with the body comdenstion water by spray pattern, then in getting into the protection casing, is blown the air to the bottle through ...

Method for measuring temperature distribution of diamond anvil cell in cryostat

According to the method for measuring the temperature distribution of the diamond anvil cell in the cryostat, a new thermal resistor and thermocouple arrangement method is adopted on the cryostat and the diamond anvil cell, experimental temperature measurement and finite element analysis are combined, and finally measurement of the sample temperature and the temperature distribution under low temperature and high pressure is achieved. Compared with a non-contact type temperature measurement method and a traditional contact type temperature measurement method, the temperature difference between the sample cavity and the clamp can be accurately obtained, the method is simple and efficient, and the calibration speed and accuracy are improved. The method is easy to implement and good in experiment repeatability; the problems that the sample cavity is small in size, the low-temperature thermostat is high in sealing requirement, and a thermocouple is not easy to arrange are solved; experimental ...

Hydrogel composite material, preparation method thereof and eye medicine

The invention relates to the technical field of medical hydrogel materials, in particular to a hydrogel composite material, a preparation method thereof and an eye medicine. The hydrogel composite material comprises hydrogel and EGCG (epigallocatechin gallate) loaded on the hydrogel; the hydrogel is prepared from methoxy polyethylene glycol amine and L-tyrosine-NCA. The hydrogel composite material provided by the invention is free of toxic and side effects and relatively long in retention time on ocular surfaces.

Generator modeling method for dynamic calculation of power system

The invention belongs to the technical field of electric power systems, and particularly relates to a generator modeling method for dynamic calculation of an electric power system, which is based on an all-pure embedded dynamic calculation operation rule and comprises the following steps of: 1, expanding an original all-pure function into a power series coefficient; 2, expanding all-pure embedding to a fourth order or a sixth order; and step 3, on the basis of the fourth-order or sixth-order model of the generator, obtaining the fourth-order or sixth-order model of the generator under full-pure embedding according to a differential transformation rule of full-pure embedding. The generator model under full-pure embedding is established based on the full-pure embedding method, so that a nonlinear algebraic equation can be converted into a linear equation, and the generator model under full-pure embedding is obtained. According to the generator modeling method based on all-pure embedding, ...

SMART SUGGESTIONS PERSONALIZATION WITH GLMIX

In an example, an indication of a plurality of different entities in a social networking service is received, including al least two entities having a different entity type. Then a plurality of user profiles in the social networking service are accessed A machine-learned model is then used to calculate, based on co-occurrence counts reflecting a number of user profiles in the plurality of user profiles in which corresponding nodes co-occurred, a similarity score between a first node and second node by computing distance between the first node and the second node in a d-dimensional space on which a plurality of entities are mapped, the similarity score generated using a generalized linear mixed model having a global coefficient vector applied to global function pertaining to the co-occurrence counts and a first random effects coefficient vector applied to a random effects per-country function. 1. A system comprising:one or more processors, retrieve an indication of a plurality of different entities in a social networking service, including at least two entities having a different entity type,', 'access a plurality of user profiles in the social networking service;', 'use a machine-learned model to calculate, based on co-occurrence counts reflecting a number of user profiles in the plurality of user profiles in which corresponding nodes co-occurred, a similarity score between a first node and second node by computing distance between the first node and the second node in a d-dimensional space on which a plurality of entities are mapped, the similarity score generated using a generalized linear mixed model having a global coefficient vector applied to global function pertaining to the co-occurrence counts and a first random effects coefficient vector applied to a random effects per-country function; and', 'use the similarity score to determine an entity to display on a computing device., 'a computer-readable medium having instructions stored thereon, which, when ...

DEVICE AND METHOD FOR ARTIFICIAL NEURAL NETWORK OPERATION

Aspects of data modification for neural networks are described herein. The aspects may include a data modifier configured to receive input data and weight values of a neural network. The data modifier may include an input data configured to modify the received input data and a weight modifier configured to modify the received weight values. The aspects may further include a computing unit configured to calculate one or more groups of output data based on the modified input data and the modifier weight values. 1. An apparatus for modifying data for neural networks , comprising: an input data modifier configured to modify the one or more groups of input data, and', 'a weight modifier configured to modify the one or more weight values; and, 'a data modifier configured to receive one or more groups of input data, one or more weight values, and one or more connection values, wherein the data modifier includesa computing unit configured to calculate one or more groups of output data based on the modified input data and the modified weight values.2. The apparatus of claim 1 ,wherein the input data modifier further includes an input data pruning module configured to modify the connection values in accordance with the one or more groups of input,wherein the input data pruning module is further configured to set at least one of the one or more connection values to zero,wherein the at least one of the one or more connection values correspond to at least one of the one or more groups of input data, andwherein absolute values of the at least one of the one or more groups of input data are less than or equal to a first threshold value.3. The apparatus of claim 2 , wherein the input data pruning module is further configured to delete at least one of the one or more groups of input data based on the modified connection values.4. The apparatus of claim 2 , wherein the input data pruning module is further configured to set each of the one or more connection values to a value equal to ...

METHOD OF TREATING HIV INFECTION BY PREVENTING INTERACTION OF CD4 AND GP120

Conversion method, apparatus, computer device, and storage medium

... 769-1949831AU A model conversion method, apparatus, computer device, and storage device are disclosed. The method an convert an initial offline model to a target offline model. The model conversion method, apparatus, computer device, and storage device may greatly reduce data input of the computer device and have a simple conversion process, and may thus reduce an amount of data processing of the computer equipment, improve rocessing efficiency and reduce power consumption.

Making method of zero-added acorn shell pigment ice cream

The invention relates to a making method of ice cream, and concretely relates to a making method of zero-added acorn shell pigment ice cream. The problems of insecurity of a common chemical pigment, and easy thawing, low expansion rate and low stability of ice cream are mainly solved in the invention. The method comprises the following steps: adding white granulated sugar into butter, whipping the butter, adding yolk in a grading manner, and uniformly stirring; heating milk in a water bath, adding starch, gelatinizing starch in warm water, and cooling; and uniformly mixing milk and the whipped butter and yolk, adding an acorn shell pigment, ageing at 4DEG C for 4h, putting the obtained mixture in an ice cream making machine, and freeze-stirring at -15 - -20DEG C for 10-30min to obtain the acorn shell pigment ice cream. The ice cream made through the method has the advantages of abundant nutrition, realization of safety and health due to use of the natural pigment, high expansion rate, difficult ...

Preparing and impregnating method of water-dilutable PU resin sizing agent for sticks buckles

The present invention relates to water-dilutable PU resin sizing agent which comprises the following components by weight: 60 to 120 shares of water-dilutable resin of which the modulus is 100 to 500 kg/cm and the solid content is 20 % to 35%, 150 to 250 shares of deionized water, and 0.1 to 0.8 shares of firming agent. The present invention also provides a preparing method of the water-dilutable PU resin sizing agent and a method of the after-treatment by using sticks buckles of the water-dilutable PU resin. The water-dilutable PU resin sizing agent has the advantages of simple preparation, low cost, reliable use and safety and source and energy saving; environment pollution and public hazard are reduced; and meanwhile, the work environment of labors is improved. The present invention completely confirms to the tendency of world environment. The sticks buckles processed by the water-dilutable PU resin sizing agent have the advantages of good hand feeling, rich of elasticity, strong solid ...

Work order distributing method and alarm information filtering device in communication network

The invention discloses a work order distribution method and an alarm information filtering device in a communication network, which are used for solving the problem of system resource waste due to the repeated distribution of the work order in the prior art. The method comprises the following steps of: acquiring alarm information meeting the work order distribution condition in the communication network by the alarm information filtering device; determining the alarm information with the same characteristic data from the acquired alarm information; configuring the work order for the determined alarm information with the same characteristic data according to a way of configuring one work order for the alarm information with the same characteristic data by the alarm information filtering device; and distributing the configured work order by the alarm information filtering device.

Nano-size incombustible polymer polyether polylol and its preparing method and application

Bandwidth regulation and control method for acousto-optic tunable filtering type spectral imager

A bandwidth regulation and control method for an acousto-optic tunable filtering type spectral imager realizes spectral bandwidth regulation and control through superposition of a plurality of spectral channel signals, and comprises the following steps: (1) determining a target spectral bandwidth, the number of superposed working wavebands and the central wavelength of an initial waveband according to actual requirements; (2) calculating the central wavelength of the working wave band participating in superposition according to the target spectral bandwidth, the number of superposed working wave bands, the central wavelength of the initial wave band and the spectral characteristics of the spectral imager; (3) determining the integral time of the first working wave band according to actual requirements; (4) calculating the integral time of other working wavebands according to the diffraction efficiency and quantum efficiency of the spectral imager and the integral time of the first working ...

APPARATUS AND METHODS FOR GENERATING DOT PRODUCT

Aspects for generating a dot product for two vectors in neural network are described herein. The aspects may include a controller unit configured to receive a vector load instruction that includes a first address of a first vector and a length of the first vector. The aspects may further include a direct memory access unit configured to retrieve the first vector from a storage device based on the first address of the first vector. Further still, the aspects may include a caching unit configured to store the first vector. 1. An apparatus for vector dot product computation in a neural network , comprising:a controller unit configured to receive a vector dot product instruction that includes a first address of a first vector and a second address of a second vector; and wherein the one or more multipliers are respectively configured to multiply, in response to the vector dot product instruction, a first element of the first vector with a corresponding second element of the second vector to generate one or more multiplication results; and', 'wherein the adder is configured to sum, in response to the vector dot product instruction, the multiplication results to generate a vector dot product computation result., 'a computation module that includes one or more multipliers and an adder,'}2. The apparatus of claim 1 , wherein the one or more multipliers are configured to transmit the multiplication results directly to the adder.3. The apparatus of claim 1 , wherein the vector dot product instruction further includes a length of the first vector.4. The apparatus of claim 3 , wherein the vector dot product instruction further includes a length of the second vector.5. The apparatus of claim 4 , further comprising a direct memory access unit configured toretrieve the first vector based on the first address and the length of the first vector, andretrieve the second vector based on the second address and the length of the second vector.6. The apparatus of claim 5 , further ...

LITHIUM-ION BATTERY ELECTROLYTE AND LITHIUM-ION BATTERY

The present invention discloses a lithium-ion battery electrolyte and a lithium-ion battery. The electrolyte comprises an organic non-aqueous solution, a lithium salt, and an additive. The additive comprises: (A) fluoroethylene carbonate; (B) at least one compound from the following: a saturated dinitrile or an unsaturated nitrile as represented by structural formula (1), wherein Ris an unsaturated hydrocarbon group with 3-6 carbon atoms and Ris an alkene group with 2-5 carbon atoms; and (C) at least one unsaturated phosphate ester as represented by structural formula (2), wherein R, R, and Rare each a hydrocarbon with 1-4 carbon atoms, and at least one of R, R, and Rcontain an unsaturated hydrocarbon with a triple bond. 2. The electrolyte according to claim 1 , wherein the additive (A) accounts for 1% to 10% of the total weight of the electrolyte.3. The electrolyte according to claim 1 , wherein the saturated dinitrile compound accounts for 1% to 5% of the total weight of the electrolyte; and the unsaturated nitrile compound represented by structural formula 1 accounts for 0.1% to 3% of the total weight of the electrolyte.4. The electrolyte according to claim 1 , wherein the additive (C) accounts for 0.1% to 2% of the total weight of the above electrolyte.6. The electrolyte according to claim 1 , wherein the unsaturated phosphate ester compound represented by structural formula 2 is tripropargyl phosphate.7. The electrolyte according to claim 1 , wherein the non-aqueous organic solvent is selected from one or more of ethylene carbonate claim 1 , propylene carbonate claim 1 , butylene carbonate claim 1 , dimethyl carbonate claim 1 , diethyl carbonate claim 1 , ethyl methyl carbonate and methyl propyl carbonate. carbonate.8. The electrolyte according to claim 1 , wherein the lithium salt is selected from one or more of LiPF claim 1 , LiBF claim 1 , LiSbF claim 1 , LiAsF claim 1 , LiN(SOCF) claim 1 , LiN(SOCF) claim 1 , LiC(SOCF)and LiN(SOF).9. A lithium-ion battery ...

Method of treating hiv infection by preventing interaction of cd4 and gp120

Novel simple and easy automatic eddy current inspection device

The utility model discloses a combination formula probe and automatic auxiliary device. Combination formula probe is main to include that one is worn formula annular probe and a miniature probe outward, is used for fixing a position the position of axial and circumference respectively. Automatic auxiliary device mainly includes workpiece holder, probe fixture, controlling means, motion, frame and detachable base. The instrument fixture include wheel and concentric fixing device, and its principle is similar to the triangle chuck. The wheel links to each other with the power structure and to let whole mechanism move on the pipeline. The probe fixture are fixed and are removed the probe. The rotation that axial displacement also can do circumference can be done to the probe fixture on the track of frame. The motion mainly includes motor and drive disk assembly. The frame includes guide rail and support body. All cyclic annular mechanisms of this equipment all adopt open design, and openable ...

Change rotation angle's mechanism

The utility model relates to a change rotation angle's mechanism, it includes: driving part and by the moving part, wherein, the driving part with by the coaxial setting of moving part, just the driving part is located by the inside of moving part, internal design has the slotted hole by the moving part, some of driving part is inlayed in this slotted hole. The utility model discloses a portion rotating when rotary motion is to the driving part to by the moving part for empty rotatory, the rotation angle that is only met the demands by the rotation angle of moving part can, reliable and simple, convenient to use, with low costs.

Method and device for obtaining Snapshot and electronic device

The embodiment of the invention provides a method and device for obtaining a Snapshot and an electronic device, and relates to the technical field of computer storage. The Snapshot can be quickly established. The method can include the steps that data collection information is calculated; according to the data collection information, the corresponding relationship between a data page number needing to be moved and a target idle page number is preserved; according to the corresponding relationship, a mapping table is backed up, and then the Snapshot is created.

Paraffin granulation underwater production method

The invention relates to a production method for paraffin wax underwater granulation, which comprises the following steps: liquid wax is sprayed out of the lower part of a granulator through nozzles of a switch board and a heated board, solidified into solid wax in demineralized water, subjected to spraying, draining and air drying, and then solidified into finished products; hot demineralized water is respectively driven into the switch board and the heated board of the granulator, and the temperature is controlled; the demineralized water is returned to a circulating tank through the granulator and a filter tank, sprayed downwards above a cooling chute as second spray water through heat exchange, partially overflows into the filter tank through a draining sieve and a cold water buffer tank and returned to the circulating tank, and partially serves as first spray water through the cold water buffer tank; the water coming from a refrigerator is returned for circulation after heat exchange ...

Method for synthesizing C14-C20 alkyl benzene for fuel scavenge by alkylation process

The invention relates to a method for synthesizing alkylbenzene for C14-C20 oil displacement for tertiary oil reclaim of an oil field by dehydrogenating and alkylating heavy liquid wax. A C14-C20 heavy liquid wax dehydrogenated alkane-alkene mixture and benzene are taken as raw materials and are subjected to HF acid alkylation reaction at a temperature of between 5 and 54 DEG C, the reaction time is between 5 and 30 minutes, the sedimentation time is between 1 and 24 hours, the mol ratio of benzene to alkene is between 4 and 6 to 1, the volume ratio of HF acid to hydrocarbon is between 1.5 and 2 to 1, a reaction product is subjected to pressure-reducing fractionation to obtain a product, the distillation range of the fractionation is between 330 and 450 DEG C, and the pressure is between 0.7 and 0.8MPa. The alkylbenzene for oil displacement ensures that the reclaiming ratio of crude oil is improved by more than 15 percent in the large-scale application of the oil field, and the interfacial ...

Solar cell module packaging adhesive film, preparation method thereof and solar cell module

The invention relates to the technical field of solar cell packaging, in particular to a solar cell module packaging adhesive film, a preparation method thereof and a solar cell module. The solar cell module packaging adhesive film comprises a first adhesive film, a water blocking film, a second adhesive film and a supporting film which are bonded in sequence. According to the packaging adhesive film for the solar cell module, the components of the adhesive films cooperate with each other, and the adhesive properties of the first adhesive film and the second adhesive film are different, so that the overall cohesiveness of the packaging adhesive film is better, the stability of the water-blocking film is improved, and the cohesiveness of the packaging adhesive film for the solar cell module is improved.

Method for training deep neural network model and related product

The invention relates to a method for training a deep neural network model and a related product. The method comprises the following steps: acquiring full-precision training parameters related to a linear layer of the deep neural network model in a forward propagation period and a back propagation period; carrying out adaptive layer-by-layer scaling quantization processing on the full-precision training parameters to obtain quantized training parameters; and in the training process of the deep neural network model, replacing a full-precision multiplication and accumulation processing operation for full-precision training parameters in each linear layer with an addition processing operation and a symbol flipping processing operation for the quantized training parameters so as to realize low-energy-consumption training of the deep neural network model. Through the scheme of the invention, the energy consumption in the deep neural network training process is effectively reduced, and the precision ...

Method and device for processing random access preamble

Disclosed are a method and device for processing a random access preamble. The method comprises: determining, on the basis of the coverage radius of a cell, a time interval time slot (GT) that is needed when a user equipment (UE) communicates with a base station, where the coverage radius of the cell is greater than 100 kilometers; and, determining, on the basis of the GT, the random access preamble transmitted by the UE to the base station, where the random access preamble is transmitted by means of a special subframe and a regular subframe in a radio frame when the UE and the base station are communicating. The present invention solves the problem that the maximum coverage range of a single-station cell cannot satisfy a need in a special scenario, further increases the coverage radius of the single-station cell, and satisfies the need in the special scenario for an extra-large radius in cell coverage range.

Method for positioning oscillation disturbance source in power system

The invention discloses a method for positioning an oscillation disturbance source in a power system. The method comprises the following steps of: acquiring voltage, current, power physical quantity of a plurality of measuring points in a multi-voltage class power grid and meeting the following conditions: 1, whether each measuring point is subjected to low-frequency synchronous power oscillationis monitored in real time, and an oscillation amplitude, a main oscillation mode frequency, main oscillation mode damping, an initial phase of the main oscillation mode and oscillation initial time information are provided; 2, two frequencies in the main oscillation mode of the measuring points which are subjected to low-frequency oscillation accord with the condition that an absolute value of delta f is less than delta Fset; 3, the oscillation initial time of the measuring points which are subjected to low-frequency oscillation has a sequence relationship, and a clear characteristic that theoscillation ...

Special high-pressure water jet washing machine for industrial mechanical parts

The invention discloses a special high-pressure water jet washing machine for industrial mechanical parts. The high-pressure water jet washing machine comprises a washing platform, wherein a high-pressure nozzle is installed in one side face of the washing platform, a moving target surface device capable of moving vertically and horizontally and moving front and back is arranged on the other opposite side of the washing platform, the moving target surface device is provided with a pressure sensor capable of measuring water flow pressure and a clamping element capable of clamping the parts needing to be washed, the high-pressure nozzle is communicated with a high-pressure jet pump through a pipeline, the high-pressure jet pump pumps water in a water tank, a backflow pipe returning to the inside of the water tank is arranged on the lower portion of the washing platform, and the high-pressure jet pump is connected with control switches on one side of the washing platform through wires. The ...

Industrial waste gas treatment ware

The utility model discloses an industrial waste gas treatment ware, which comprises a housin, the casing has first cavity and second cavity, and the top of first cavity and second cavity do not is provided with a spray set and the 2nd spray set, is fixed with the fan on the outer wall of the first cavity of casing, and the lower extreme of fan is provided with the fluviograph. The utility model discloses a spray the first time of first cavity, and can dispose in the waste gas can water -soluble harmful substance, and spray the second time through the second cavity, and residual gas obtains furtherly and handles, and through twice processing, the pollutant in the waste gas can obtain abundant filtration, causes the pollution to the environment in avoiding discharging the air.

Detection method and system of membrane electrode production line, terminal equipment and storage medium

The invention is suitable for the technical field of intelligent control, and provides a detection method and system of a membrane electrode production line, terminal equipment and a storage medium. The method is applied to the membrane electrode production line, and the membrane electrode production line comprises multiple pieces of production equipment. Comprising the steps of obtaining a first actual size value fed back by first equipment for a target material and a second actual size value fed back by second equipment for the target material; generating a deviation fluctuation value according to the first actual size value and the second actual size value; comparing the deviation fluctuation value with a preset deviation threshold value; and if the deviation fluctuation value is greater than or equal to the deviation threshold value, marking the first equipment and the second equipment as key detection equipment. According to the method and the device, production personnel can conveniently ...

Energy storage system generator tripping method, system and equipment and storage medium

The embodiment of the invention discloses an energy storage system generator tripping method, system and device and a storage medium, and the method comprises the steps: obtaining an electromagnetic power curve acceleration area and an electromagnetic power curve deceleration area through employing a first electromagnetic power curve, a second electromagnetic power curve and third equivalent mechanical power; and obtaining the generator tripping amount of the energy storage system by using the electromagnetic power curve acceleration area and the electromagnetic power curve deceleration area. The acceleration area of the electromagnetic power curve can more accurately represent the electromagnetic power change caused after the energy storage system is locked, and the deceleration area of the electromagnetic power curve can more accurately represent the electromagnetic power change caused after the energy storage system is put into the DC line. More accurate generator tripping amount can ...

NEURAL NETWORK COMPUTING METHOD, SYSTEM AND DEVICE THEREFOR

The present disclosure provides a neural network computing method, system and device therefor to be applied in the technical field of computers. The computing method comprises the following steps: A. dividing a neural network into a plurality of subnetworks having consistent internal data characteristics; B. computing each of the subnetworks to obtain a first computation result for each subnetwork; and C. computing a total computation result of the neural network on the basis of the first computation result of each subnetwork. By means of the method, the present disclosure improves the computing efficiency of the neutral network. 1. A neural network computing method , comprising the following steps:A. dividing a neural network into a plurality of subnetworks having consistent internal data characteristics;B. computing each of the subnetworks to obtain a first computation result for each subnetwork; andC. computing a total computation result of the neural network on the basis of the first computation result of each subnetwork.2. The computing method according to claim 1 , wherein the step A comprises:A1. dividing the neural network into a plurality of subnetworks having consistent internal data characteristics on the basis of output neurons of the neural network;A2. dividing the neural network into a plurality of subnetworks having consistent internal data characteristics on the basis of input neurons of the neural network; andA3. dividing the neural network into a plurality of subnetworks having consistent internal data characteristics on the basis of neuron weights of the neural network.3. The computing method according to claim 2 , wherein the step A3 comprises:dividing the neural network into a plurality of subnetworks having consistent internal data characteristics on the basis of distribution of the neuron weights of the neural network; ordividing the neural network into a plurality of subnetworks having consistent internal data characteristics on the basis of ...

Hard and high-calcium oyster mushroom cultivation base stock and preparation method thereof

The invention relates to hard and high-calcium oyster mushroom cultivation base stock and a preparation method thereof. The hard and high-calcium oyster mushroom cultivation base stock comprises the following components: corncob, bean dreg, clam shell, hydrochloric acid solution, lime powder, light calcium carbonate and zinc sulfate. According to the invention, calm shell powder is added into the cultivation base stock, and decomposed into components which are easy for being utilized by virtue of piling-covering and fermenting processes; and meanwhile, a great deal of the calm shell powder is utilized, so that oyster mushroom hyphae absorb calcium element as much as possible, and therefore, calcium element content in the mushroom body is increased, the mushroom body is hard, stiff and smooth in texture, and outline commodity and texture mouthfeel as well as storage and transportation resistance are obviously improved. Besides, the calm shell is always the waste which cannot be utilized ...

Detection method and system for double-roller heat transfer printing equipment

The invention discloses a detection method and system for double-roller heat transfer printing equipment, and relates to the technical field of fuel cells, and the detection method comprises the steps: carrying out defect position detection on a film material before transfer printing through a first NG identifier identification assembly, and obtaining film material NG data before transfer printing; performing width detection on the transferred film material through a width detection assembly to obtain film material transfer width data; defect position detection is conducted on the transferred film material through a second NG marking assembly, and transferred finished product NG data are obtained; carrying out margin detection on the transferred waste through a margin detection assembly to obtain transfer margin data; and evaluating the transfer printing quality of the thermal transfer printing equipment in combination with the film material NG data, the film material transfer printing ...

Group cooperation task-oriented information transmission method

The invention discloses an information transmission method for a group cooperative task, and belongs to the field of command and control. The method mainly comprises the following steps of: dividing message types required by a group cooperation task, wherein the message types comprise role type messages, situation type messages, abnormal state type messages, task related messages and other messages; defining 22 groups of group collaborative tasks and organization forms thereof; designing two message transmission modes, namely a centralized mode with a stable communication link and a centerless mode with an unstable communication link, oriented to the field of group cooperative task command and control; and designing a message sending and handshaking mechanism in a centerless mode oriented to the field of group collaborative task command and control. According to the cooperative task message transmission mechanism, the problem that group cooperative task information transmission is difficult ...

HELIX FERROELECTRIC LIQUID CRYSTAL CELL WITH POLYMER-STABILIZED AND ELECTRICALLY SUPPRESSED PROPERTIES

The present invention provides an electrically suppressed helix ferroelectric liquid crystal (ESHFLC) cell with polymer stabilization. The cell has a liquid crystal (LC) material that is a mixture comprising a monomer, a photo-initiator and a ferroelectric liquid crystal (FLC) where a polymer network has been established at a certain temperature to achieve constraints of ESHFLC electro-optical mode. The resultant mixture is characterized by a helix pitch less than and comparable to a FLC layer thickness of the cell or below the thickness, and provides a selective reflection in an UV region. The concentration of the monomer in the pure FLC mixture has also been optimized for the phase diagram, scattering and the tilt angle. The resultant mixture, i.e. the polymer-stabilized ESHFLC cell, follows all constraints of the ESHFLC electro-optical mode and presnets electro-optical properties similar to a typical ESHFLC cell. COPYRIGHT KIPO 2015 ...

Dianthus caryophyllus seedling plug tray cottage and cultivation method

The invention relates to a dianthus caryophyllus seedling plug tray cottage and cultivation method, and belongs to the technical field of flower cultivation. The method includes the steps of female parent seedling selecting, seedling collecting, ear cutting, seedling processing, substrate preparing, plug tray preparing, cottage mode selecting, management after cottage, finished product seedling preparing and the like. By means of the method, root cluster performance of root systems of dianthus caryophyllus seedlings is strong, influences due to environment changes are few, root emergence speed is high, seedling emergence speed is high, the seedlings are strong, the survival rate can reach 99.4%, the phenomenon that the root systems are damaged due to seedling lifting after seedling establishment can be avoided, meanwhile the substrate which is used in the method can be used repeatedly, substrates are saved, and cost is reduced. Besides, a plug tray is adopted for modularized cottage and ...

Method for implementing uniform transmission of short message in entire network

The invention discloses a method for implementing uniform transmission of a short message in entire network, which is performed according to the following steps of: a, setting a master server as a uniform interface for processing the short message, and respectively designing regular expressions of mobile phone numbers of three operators including China Mobile, China Unicom and China Telecom, wherein the regular expressions are used for identifying the operators of the mobile phone numbers; then, automatically identifying and dividing the mobile phone numbers, which are input while transmitting the short message, into three mobile phone number groups of the China Mobile, the China Unicom and the China Telecom based on the designed regular expressions; respectively transmitting the values of the identified and divided mobile phone number groups and the content to be transmitted into corresponding telecom operator short message interfaces; if the mobile phone number does not belong to the ...

Storage method of index data and storage control device

The invention discloses a storage method of index data and a storage control device; the method comprises the following steps: index data of a disk of a coder, which is needed to be stored to a storage device, is cached; when the quantity of the cached index data reaches preset threshold, or a trigger event is detected, the cached index data is stored to the disk. By adopting the invention, the efficiency on reading-in the index data and video data to the disk can be improved.

Matlab-based RTS noise detection analysis system for image sensor

The invention relates to a Matlab-based RTS noise detection analysis system for an image sensor. The system is characterized by using Matlab to read collected images and converting the images as pixel signals of a pixel gray value; performing white noise filtering to each pixel signal to obtain the pixel signals; using a gradient detecting method to obtain gradient change signals of the pixel signals; obtaining a pixel gradient signaling noise threshold via a threshold algorithm, and then denoising the gradient signals; counting number of gradient pulse of each pixel and outputting as an RTS noise detection result; averaging the signals between each two pulses of a pixel gradient signal to obtain an average dark current base of the pixel RTS noise; obtaining a reconstructed RTS noise signal by combining the average dark current base of the RTS noise and a gradient pulse signal. The system of the invention can be applied to any visible image sensors and can automatically realize RTS noise ...

Optic fibre printing opacity concrete display device

The utility model discloses an optic fibre printing opacity concrete display device contains the printing opacity building block, the printing opacity building block embeds a plurality of optic fibre, and the light source is connected to optic fibre one end, and a plurality of optic fibre can be the matrix or a certain specific pattern is arranged, are connected with the light source cable on the light source, a plurality of light source cables all are connected with the controller, optic fibre can mix the optical fiber bundle that forms for at least 2 kinds different colour optic fibre, the light source can be for the light source group of built -in 2 kind at least different colours, the printing opacity building block can for printing opacity concrete, opaque syrup, steel construction, plastics, wooden one of them make, the utility model discloses can be simply, low -cost and low power consuming ground realize daytime indoor lighting and indoor LOGO shows, night outdoor LOGO show that ...

Production process of waterborne polyurethane coating material having high tension force coating film

The present invention discloses a production process of a waterborne polyurethane coating material having a high tension force coating film. According to the production process, polyether, polyester polyol and isocyanate are adopted as raw materials, a hydrophilic chain extender is gradually added during the reaction process, an ethylene glycol dimethyl ether-acetone mixed solvent or diethylene glycol dimethyl ether-acetone mixed solvent is added when the reaction achieves the theoretical NCO% value, and finally salification, emulsifying and chain extending are performed to obtain the product. According to the present invention, the hydrophilic groups in the obtained product of the present invention are broadly distributed in the molecular chain compared with the product obtained through the traditional process, wherein the hydrophilic groups are not only distributed in the prepolymer chain end and also distributed in the chain so as to further improve the hydrophilic efficiency of the ...

All-digital simulation system and method for infrared imaging system

The invention discloses an all-digital simulation system and method for an infrared imaging system. The simulation system is composed of a simulation platform, a simulation member database, a simulation information table, a plurality of simulation members and a simulation client, wherein the simulation platform is used for comprehensively managing an overall simulation process; the simulation member database is used for storing simulation member information; the simulation information table is used for storing simulation configuration information; the simulation members integrate one or more simulation models and interact with the simulation platform by virtue of the simulation client. The simulation method comprises the following specific steps: building and registering the simulation members; configuring a simulation process; carrying out simulation propulsion on the simulation members by the simulation platform according to the simulation process; and after the simulation is completed ...

Method for accelerating germination in beer barley malting

The invention discloses a method for accelerating germination in beer barley malting by combining brassinolide (BR), 6-benzyladenine (6-BA) and gibberellic acid (GA), so as to change a process for accelerating barley germination by singly using the gibberellic acid during the prior malting, reduce the amount of the gibberellic acid, improve the germination ratio of the barley, shorten the germination time, and avoid increased hidden trouble of chromaticity of malts caused by using a large amount of the gibberellic acid; and the property of the brassinolide is utilized to improve the mixed-bacteria resistant performance of green malts so as to finally achieve the aims of improving the quality of the malts and reducing the production cost. The method comprises processes of grading, soaking the malts and germinating, and is applied in different periods of the barley germination process through using the brassinolide (BR) with extremely low concentration, efficiency and natural plant hormone ...

Three-dimensional vehicle tracking method and device based on monocular vision and electronic equipment

The invention relates to the technical field of target three-dimensional tracking detection, and provides a three-dimensional vehicle tracking method and device based on monocular vision and electronic equipment. The method comprises the following steps: acquiring a surrounding field image of a vehicle, wherein the surrounding field image is a two-dimensional image and comprises at least one target vehicle; inputting the two-dimensional image into a pre-trained deep learning neural network model to obtain two-dimensional composite information of the target vehicle in the two-dimensional image; based on the parameters of the camera and the two-dimensional composite information of the target vehicle, determining that the vehicle is three-dimensional ranging information of the target vehicle under the vehicle body coordinate system; and determining a three-dimensional tracking state of the target vehicle based on the re-identification feature, the two-dimensional composite information and ...

Hybrid simulation system and hybrid simulation method suitable for stability control device interface

The invention provides a hybrid simulation system suitable for a stability control device interface. The hybrid simulation system includes an RTDS simulation machine, a high-speed parallel computer, an electromagnetic I/O module, an electromechanical I/O module and an AC/DC stability control device; the RTDS simulation machine is connected with the high-speed parallel computer through an optical fiber; the RTDS simulation machine is connected with the AC/DC stability control device through the electromagnetic I/O module; one end of the electromagnetic I/O module is connected with the RTDS simulation machine through an optical fiber; the other end of the electromagnetic I/O module is connected with the AC/DC stability control device through a cable; the high-speed parallel computer is connected with the AC/DC stability control device through the electromechanical I/O module; one end of the electromechanical I/O module is connected with the high-speed parallel computer through an optical ...

Synthesize PFT and detect training ware

The utility model provides a synthesize PFT and detect training ware, including respiratory tube, air cavity, electronic valve and the blast pipe that connects gradually, is furnished with pressure sensor on the air cavity, electronic valve and pressure sensor all are connected with microcontroller, last power, control button, display screen and the external interface of still being connected with of microcontroller. The utility model discloses has following advantage: it can the pressure detection function not only to contain the vital capacity and breathe flesh to contain and breathe that flesh can train and function such as row's phlegm training, carry out the operating guideline through the information feedback of display screen with electron voice prompt in the use, the person of facilitating the use carries out the independent operation, reduces medical personnel's work load.

Multichannel soil temperature takes notes appearance

The utility model provides a problem that multichannel soil temperature takes notes appearance belongs to soil parameters and detects technical field, and regularly the volume of fixing a point at surveying work is big, the degree of difficulty is big in the manual work that aim at solution prior art exists. The utility model discloses a multichannel soil temperature takes notes appearance includes power, singlechip, multichannel temperature data collection unit, temperature memory cell and clock module, the power respectively with the power supply is connected to singlechip, multichannel temperature data collection unit, temperature memory cell and clock module, the singlechip is connected through signal transmission line with multichannel temperature data collection unit, temperature memory cell and clock module respectively, multichannel temperature data collection unit is inputed measuring temperature data to the singlechip in, the clock module with the time data input extremely in ...

Matrix for dianthus caryophylus seed and seedling tray cutting and breeding

The invention discloses a matrix for dianthus caryophylus seed and seedling tray cutting and breeding, and belongs to the technical field of flower culture. The matrix is characterized by being prepared by uniformly mixing 30-50wt% of perlite, 25-35wt% of vermiculite and 25-35wt% of grass carbon. The matrix can provide sufficient nutritional ingredients for seedling raising of dianthus caryophylus, is high in rooting speed, more in secondary branch, and healthy and strong in seed and seedling; the matrix applied in tray cutting is high in wrapping performance; root systems of dianthus caryophylus seeds and seedlings are high in root wrapping performance, small in environmental change influence, high in survival rate, and high in seedling survival speed; and root system damage due to seedling lifting after seedling maturity can be avoided.

Model predictive control frequency modulation method

The embodiment of the invention discloses a model predictive control frequency modulation method, and the method comprises the steps: designing virtual inertia control in a preset power system frequency modulation model of an ith region, and compensating a virtual inertia control instruction outputted by the virtual inertia control to a model predictive control model of the ith region, obtaining a target model predictive control model of the ith area, discretizing the target model predictive control model of the ith area, then solving the discretized target model predictive control model to obtain an optimal control sequence, taking a first element in the optimal control sequence as a control instruction of the current sampling moment, repeating the solving steps to obtain a control instruction of each sampling moment, and obtaining a target model predictive control model of the ith area; the control instruction is used for controlling the power system of the ith area to perform frequency ...

Non-electric cutting knife capable of being quickly disassembled and assembled

The non-electric cutting knife comprises a medicine cylinder and an excitation handle, the excitation handle comprises a handle outer shell and a handle body, the handle outer shell comprises an upper shell and a lower shell, the handle body is installed in the handle outer shell, the medicine cylinder comprises a medicine cylinder base and a medicine cylinder body which are integrally formed, and the medicine cylinder base and the medicine cylinder body are arranged in the medicine cylinder base. The cartridge base is in contact fit with the handle body in the handle shell, the cartridge barrel extends out of the handle shell, and a locking mechanism for fixing the upper shell and the lower shell is arranged at the end, away from the cartridge, of the handle shell. By means of the locking mechanism, the upper shell and the lower shell can be rapidly and conveniently disassembled or matched, so that a medicine cylinder of the non-electric cutting knife is very convenient to disassemble ...

Finishing method of regenerated cellulose fiber fabric

The invention relates to the technical field of textile, in particular to a finishing method of regenerated cellulose fiber fabric. Warp yarns and weft yarns are regenerated cellulose fiber yarns, and the processing procedure of the regenerated cellulose fiber fabric sequentially comprises singeing, desizing, pretreatment, liquid ammonia treatment, tentering and preshrinking. According to the method disclosed by the invention, the damage of the regenerated cellulose fiber fabric in liquid ammonia processing is reduced to the greatest extent through sufficient oil removal of desizing, a pre-stentering step before liquid ammonia adding and adjustment on a liquid ammonia process, so that the problems of oil stains and hard hand feeling in the liquid ammonia processing process of the fabric are solved; production and processing of liquid ammonia finishing of the regenerated cellulose fibers are smoothly achieved, and the soft, smooth and comfortable regenerated cellulose fiber high-quality ...

HAND-HELD MOBILE APPARATUS AND METHOD OF QUCKLY DISPLAYING ONLINE COMMUNITY FRIENDS' IMAGES APPLICABLE THERETO

A hand-held mobile apparatus and method of quickly displaying online community friends' images applicable thereto are disclosed. The hand-held mobile apparatus includes a network communication module for linking to an online community server; a comparison module for comparing and determining whether data of a web image in the online community server matches data of a local image in a storage unit of the apparatus; a recognition module for determining an image type of the web image when the web image data is different from the local image data; a processing unit for downloading the web image of a first type in a smallest image size or the web image of a second type in an image size matching a screen resolution of the apparatus, and storing the web image in the storage unit as the local image; and a display unit for displaying the local image. 1. A hand-held mobile apparatus , comprising:a network communication module for linking to an online community server via a network system; and the online community server having a web image stored therein;a storage unit for storing a local image therein;a comparison module for comparing data of the web image with data of the local image and determining whether the web image data matches the local image data;a recognition module being connected to the comparison module for receiving the web image when the comparison module determines the web image data is different from the local image data, and for determining whether the web image is of a first image type or a second image type;a processing unit being connected to the network communication module, the storage unit, the comparison module and the recognition module; in the case the recognition module determining that the web image is of the first image type, the processing unit downloading from the online community server via the network communication module the web image in a smallest possible image size thereof; or in the case the recognition module determining that the web ...

HAND-HELD MOBILE APPARATUS CAPABLE OF QUICKLY DISPLAYING PICTURES AND METHOD OF QUICKLY DISPLAYING PICTURES APPLICABLE THERETO

A hand-held mobile apparatus and a method capable of quickly displaying picture are disclosed. The method includes the steps of linking a communication unit to a server via a network system; storing a picture downloaded from the server in a storage unit; compressing the downloaded picture with an encoder unit, so that the compressed picture has a file size smaller than the downloaded picture; using a recognition unit to recognize whether the compressed picture is a head image; if yes, using a naming unit to specify a first file name to the compressed picture; storing the compressed picture in the storage unit; and displaying the compressed picture in a display screen. Further, a data buffer is provided to temporarily store the compressed picture transmitted by a control unit to the display screen, so that the compressed pictures can be quickly displayed on the hand-held mobile apparatus having the above units. 1. A hand-held mobile apparatus capable of quickly displaying pictures , comprising:a communication unit linking to a server via a network system; the server having at least one picture stored therein;a storage unit storing at least one picture downloaded from the server;an encoder unit compressing the downloaded at least one picture into a compressed picture, such that the compressed picture has a file size smaller than that of the downloaded at least one picture;a recognition unit recognizing whether the compressed picture matches a path characteristic or not;a naming unit specifying a first file name to the compressed picture, provided the recognition unit recognizes that the compressed picture matches the path characteristic;a display screen displaying the compressed picture; anda control unit being electrically connected to the communication unit, the encoder unit, the recognition unit, the naming unit, the storage unit and the display screen for controlling the units connected thereto to work, storing the compressed picture in the storage unit, and ...

MANAGING UNIQUELY TAGGED ITEMS USING THE INTERNET

The invention teaches managing an item in the Internet of Things, wherein the system comprises: an item registration module, configured to receive at least one registration information of the item, wherein the registration information of the item includes the item's unique number marked by an information sensing device and a tag of an external data source where related information of the item is located, the at least one registration information is from at least one external data source, the unique number is the same in the at least one registration information; a storage module, configured to store the item's registration information; and a query processing module, configured to receive a query to the related information of the item, and obtain the related information of the item from the external data source which corresponds to the external data source tag included in the stored registration information of the item. 1. A system for managing an item in the Internet of Things , comprising:an item registration module, configured to receive at least one registration information of the item, wherein the registration information of the item includes the item's unique number marked by an information sensing device and a tag of an external data source where related information of the item is located, the at least one registration information is from at least one external data source, the unique number is the same in the at least one registration information;a storage module, configured to store the item's registration information; anda query processing module, configured to receive a query to the related information of the item, and obtain the related information of the item from the external data source which corresponds to the external data source tag included in the stored registration information of the item.2. The system as claimed in claim 1 , further comprising an authentication module claim 1 , configured to authenticate an authenticity of the external data ...

MULTI DIRECTION SWITCH

A multi direction switch () includes an upper housing (), a lower housing (), a bottom cover (), an elastic ring (), an actuator (), an operating button (), a terminal plate (), a moveable electrode plate (), an insulative layer (), and a fixed electrode plate (). The upper housing and the lower housing define a channel () and a first receiving space () for receiving the elastic ring. The bottom cover is attached to the lower housing for defining a second receiving space () therebetween. The actuator has a bracket () connecting with the elastic ring via the channel and an actuating portion () extending from the bracket towards the bottom cover. The terminal plate, the moveable electrode plate, the insulative layer, and the fixed electrode plate are positioned along an upper-and-lower direction and received in the second receiving space. 1. A multi direction switch comprising:an upper housing and a lower housing together defining a channel and a first receiving space communicating with the channel;a bottom cover attached to the lower housing for defining a second receiving space therebetween;an elastic ring received in the first receiving space;an actuator having a bracket extending through the channel to contact with the elastic ring and an actuating portion extending from the bracket towards the bottom cover;an operating button having an operating portion positioned above the upper housing and a protrusion extending downwardly from the operating portion, the protrusion retained by the bracket; anda terminal plate, a moveable electrode plate, an insulative layer, and a fixed electrode plate positioned in order from top to bottom and received in the second receiving space; whereinthe moveable electrode plate is mounted to the actuating portion of the actuator.2. The multi direction switch as claimed in claim 1 , wherein the moveable electrode plate is moveable along a horizontal direction.3. The multi direction switch as claimed in claim 2 , wherein the terminal ...

METHODS, COMPUTER PROGAM PRODUCTS, AND SYSTEMS FOR INCREASING INTEREST IN A MASSIVELY MULTIPLAYER ONLINE GAME

One aspect of the invention provides a method of increasing interest in a massively multiplayer online game. The method includes providing the massively multiplayer online game on each of a plurality of shards and associating a list of accomplishments with each of the plurality of shards. Another aspect of the invention provides a computer program product comprising computer-usable medium having control logic stored therein for causing a computer to implement a method of increasing interest in a massively multiplayer online game. The control logic includes: first computer readable program code for causing the computer to provide the massively multiplayer online game on each of a plurality of shards and second computer readable program code for causing the computer to associate a list of accomplishments with each of the plurality of shards. 1. A method of increasing interest in a massively multiplayer online game , the method comprising:providing the massively multiplayer online game on each of a plurality of shards; andassociating a list of accomplishments with each of the plurality of shards.2. The method of claim 1 , further comprising:publishing the list of accomplishments.3. The method of claim 2 , wherein the list of accomplishments is published via an XML feed.4. The method of claim 1 , further comprising:providing the massively multiplayer online game on a new shard;associating a list of accomplishments with each of the plurality of shards; andannouncing the time at which the new shard will be accessible.5. The method of claim 1 , wherein the accomplishments include temporal accomplishments.6. The method of claim 5 , wherein the temporal accomplishments include one or more selected from the group consisting of: the first n players to achieve a certain score claim 5 , the first n players to achieve a population claim 5 , the first n players to achieve a certain number of cities claim 5 , and the first n players to achieve a certain title claim 5 , wherein n is ...

METHODS, COMPUTER PROGRAM PRODUCTS, AND SYSTEMS FOR AWARDING ITEMS IN A MULTIPLAYER ONLINE GAME

One aspect of the invention provides a method of awarding items in a multiplayer online game. The method includes: presenting a primary award interface to a first user, the award interface configured to implement a primary award algorithm that determines whether the first user will receive an item and, if so, informs the first user of the award; and informing other users of the award to the first user. Another aspect of the invention provides a system awarding items in a multiplayer online game. The system includes a presentation module configured to present an award interface to a first user, the award interface configured implement a primary award algorithm that determines whether the first user will receive an item and, if so, informs the first user of the award; and a notification module configured to inform other users of the award to the first user. 1. A method of awarding items in a multiplayer online game , the method comprising:presenting a primary award interface to a first user, the award interface configured to implement a primary award algorithm that determines whether the first user will receive an item and, if so, informs the first user of the award; andinforming other users of the award to the first user.2. The method of claim 1 , wherein the item is randomly selected.3. The method of claim 1 , wherein the item is semi-randomly selected.4. The method of claim 1 , wherein the item is selected by the first user.5. The method of claim 1 , wherein the item is selected through a sub-game.6. The method of claim 1 , wherein the primary award interface:displays a plurality of items;sequentially highlights one or more of the plurality of items; andindicates which item will be awarded to the first user.7. The method of claim 1 , wherein the step of informing other users of the award to the first user includes sending a message to the other users.8. The method of claim 7 , wherein the message is displayed on a further interface for the multiplayer online game.9 ...

METHODS, COMPUTER-PROGAM PRODUCTS, AND SYSTEMS FOR IMPLEMENTING A COLONIAL RELATIONSHIP IN A MULTIPLAYER ONLINE GAME

One aspect of the invention provides a method for implementing a colonial relationship in a multiplayer online game. The method includes: receiving instructions from a first entity to attack a second entity; determining whether the first entity's attack is successful; and if the first entity's attack is successful, establishing a colonial relationship between the second entity and the first entity, wherein a portion of production of the second entity is transferred to the first entity. 1. A computer-implemented method for implementing a colonial relationship in a multiplayer online game , the method comprising:receiving instructions on a computer from a first entity to attack a second entity;determining on the computer whether the first entity's attack is successful; andif the first entity's attack is successful, establishing a colonial relationship between the second entity and the first entity, wherein a portion of production of the second entity is transferred to the first entity.2. The method of claim 1 , further comprising:receiving instructions from the second entity to institute an uprising; anddetermining whether the second entity's uprising is successful; andif the uprising is successful, terminating the colonial relationship.3. The method of claim 1 , further comprising:receiving instructions from the first entity to terminate the colonial relationship; andterminating the colonial relationship.4. The method of claim 1 , further comprising:receiving instructions from a third entity to attack the second entity;determining whether the third entity's attack is successful; andif the third entity's attack is successful, establishing a colonial relationship between the second entity and the third entity, wherein a portion of production of the second entity is transferred to the third entity.5. The method of claim 1 , further comprising:receiving instructions from a fourth entity to attack the first entity;determining whether the fourth entity's attack is ...

VIEWING AND DEBUGGING HDL DESIGNS HAVING SYSTEMVERILOG INTERFACE CONSTRUCTS

Methods and systems for viewing and debugging HDL designs having SystemVerilog interface constructs are provided. An HDL design code is received, wherein the design code comprises a first module, a second module and a SystemVerilog interface construct. A first object corresponding to the first module, a second object corresponding to the second module and an interface object corresponding to the interface construct are displayed in a schematic view. The interface object is disposed between the first and second objects, and a shape of the interface object is different from that of the first and second objects. The interface signals from the first object to the second object and the interface signals from the second object to the first object pass through the interface object. 1. A method for viewing and debugging an HDL design having SystemVerilog interface constructs , comprising:receiving an HDL design code, wherein the design code comprises a first module, a second module and a SystemVerilog interface construct; anddisplaying a first object corresponding to the first module, a second object corresponding to the second module and an interface object corresponding to the SystemVerilog interface construct in a schematic view,wherein the interface object is disposed between the first and second objects, and the shape of the interface object is different from that of the first and second objects,wherein interface signals from the first object to the second object and interface signals from the second object to the first object pass through the interface object.2. The method as claimed in claim 1 , wherein the shape of the interface object is a double arrow claim 1 , and each shape of the first and second objects is a rectangle.3. The method as claimed in claim 1 , wherein the first object has a first module port coupled to a second module port of the interface object for transferring of the interface signals between the first and second objects claim 1 , and the second ...

Power Socket Connector And Conductive Terminal Thereof

The present invention discloses a power socket connector, including an insulating body, a shielding case body outwardly covering the insulating body, and a terminal set installed in the insulating body. The terminal set includes a central contact terminal and a conductive terminal disposed on two sides of the central contact terminal. The conductive terminal includes a vertical flexible contact portion and a horizontal flexible contact portion, and the vertical flexible contact portion and the horizontal flexible contact portion are connected to an L-shaped holding portion via a vertical connecting portion and a horizontal connecting portion, respectively. When a butt joint power plug is inserted into the power socket connector, the vertical/horizontal flexible contact portions of the conductive terminals are simultaneously abutted against the butt joint power plug to increase the high current crossing area, thereby increasing the current conductive stability and avoiding high temperature burnout of the connector.

DISPLAY METHOD, DISPLAY APPARATUS AND ELECTRONIC DEVICE

A display method, a display apparatus and an electronic device are described. The display method is used in an electronic device. The display method includes displaying a first application-interface, wherein, first identifiers of at least one first application are displayed in a first display effect in the first application-interface, and the at least one first application is stored on server-side; receiving a selection input from a user; determining a target identifier in the first identifiers of the at least one first application according to the selection input; displaying a second application-interface, wherein, second identifiers of at least one second application are displayed in a second display effect in the second application-interface, and the at least one second application is stored on electronic-device-side; displaying the target identifier and the transformation process of the target identifier from the first display effect to the second display effect in the second application-interface. 1. A display method used in an electronic device , the display method comprising:displaying a first application-interface, wherein, first identifiers of at least one first application are displayed in a first display effect in the first application-interface, and the at least one first application is stored on server-side;receiving a selection input from a user;determining a target identifier in the first identifiers of the at least one first application according to the selection input;displaying a second application-interface, wherein, second identifiers of at least one second application are displayed in a second display effect in the second application-interface, and the at least one second application is stored on electronic-device-side;displaying the target identifier and the transformation process of the target identifier from the first display effect to the second display effect in the second application-interface.2. The display method according to claim 1 , ...

USER QUESTION PROCESSING METHOD AND SYSTEM

Provided is a user question processing method and system. The method includes: extracting first feature information from a user question; calculating the similarity between the first feature and second feature information of each of at least two websites; posting the question on at least one of the at least two websites according to the similarity. The solution of the embodiment can be applied to a website providing a question and answer service, and the website can post a received user question on another website related to the concerned field of the question, thereby enlarging the scope of the user information exchange. 1. A user question processing method , comprising:extracting first feature information from a user question;calculating a similarity between the first feature information and second feature information of each of at least two websites; and [ 'extracting textual feature information of the user question;', 'wherein the step of extracting first feature information from a user question comprises, 'the step of calculating a similarity between the first feature information and second feature information of each of at least two websites comprises:', 'obtaining content feature information of the website;', 'obtaining a first module of the user question according to the textual feature information; the first module comprises a probability of the user question belonging to each classification of at least two content classifications;', 'obtaining a second module of the website according to the content feature information; the second module comprises a probability of the website belonging to each classification of the at least two content classifications;', 'calculating the similarity between the first module and the second module, setting the similarity between the first module and the second module as a similarity between the first feature information and the second feature information., 'posting the user question on at least one of the at least two websites ...

GENERATING CANDIDATES FOR SEARCH USING SCORING/RETRIEVAL ARCHITECTURE

Techniques for generating candidates for search using a scoring and retrieval architecture and deep semantic features are disclosed herein. In some embodiments, a computer system generates a profile vector representation for user profiles based profile data, stores the profile vector representations, receives a query subsequent to the storing of the profile vector representations, generates a query vector representation for the query, retrieves the stored profile vector representations of the user profiles based on the receiving of the query, generates a corresponding score for pairings of the user profiles and the query based on a determined level of similarity between the profile vector representation of the user profiles and the query vector representation, and causes an indication of at least a portion of the user profiles to be displayed as search results for the query based on the generated scores of the user profiles. 1. A computer-implemented method comprising:for each one of a plurality of user profiles stored on a database of an online service, retrieving, by a first neural network, profile data of the one of the plurality of user profiles from the database of the online service;for each one of the plurality of user profiles, generating, by the first neural network, a profile vector representation based on the retrieved profile data of the one of the plurality of user profiles;storing the profile vector representations of the plurality of user profiles in the database of the online service;receiving, by a computer system having a memory and at least one hardware processor, a query from a computing device of a querying user subsequent to the storing of the profile vector representations, the query comprising query data, the query data comprising at least one of query text or facet selection data;generating, by a second neural network distinct from the first neural network, a query vector representation for the query based on the query data of the query in ...

GENERATING SUPERVISED EMBEDDING REPRESENTATIONS FOR SEARCH

Techniques for generating supervised embedding representations for search are disclosed herein. In some embodiments, a computer system receives training data comprising query representations including an entity included in a corresponding search query submitted by a querying user, search result representations for each one of the query representations, and user actions for each one of the query representations, and generates a corresponding embedding vector for each one of the at least one entity using a supervised learning algorithm and the received training data. In some example embodiments, the corresponding search result representations for each one of the query representations represents a plurality of candidate users displayed in response to the search queries based on profile data of the candidate users, and the user actions comprise actions by the querying user directed towards at least one candidate user in the search results. 1. A computer-implemented method comprising:receiving, by a computer system having a memory and at least one hardware processor, training data comprising a plurality of query representations, a plurality of search result representations for each one of the plurality of query representations, and a plurality of user actions for each one of the plurality of query representations, each one of the plurality of query representations representing at least one entity included in a corresponding search query submitted by a querying user, the corresponding plurality of search result representations for each one of the plurality of query representations representing a plurality of candidate users displayed in response to the plurality of search queries based on profile data of the plurality of candidate users stored on a database of an online service, the plurality of user actions comprising actions by the querying user directed towards at least one candidate user of the plurality of search results for the corresponding search query;generating, ...

GENERATING SUPERVISED EMBEDDINGS USING UNSUPERVISED EMBEDDINGS

Techniques for generating supervised embedding representations using unsupervised embedding representations and deep semantic structured models for search are disclosed herein. In some embodiments, a computer system generates a graph data structure based on accessed profile data, generates an initial embedding vector using an unsupervised machine learning algorithm, receiving training data comprising query representations, search result representations, and user actions, with each one of the plurality of query representations comprising the initial embedding vector, and generates a final embedding vector using a supervised learning algorithm and the received training data. 1. A computer-implemented method comprising:for each one of a plurality of reference users of an online service, accessing, by a computer system having a memory and at least one hardware processor, corresponding profile data of the reference user stored on a database of the online service, the accessed profile data of each reference user indicating at least one entity of a first facet type associated with the reference user;generating, by the computer system, a graph data structure based on the accessed profile data, the generated graph data structure comprising a plurality of nodes and a plurality of edges, each one of the plurality of nodes corresponding to a different entity indicated by the accessed profile data, and each one of the plurality of edges directly connecting a different pair of the plurality of nodes and indicating a number of the plurality of reference users whose profile data indicates both entities of the pair of nodes that are directly connected by the edge;generating, by the computer system, a corresponding initial embedding vector for each one of the entities indicated by the accessed profile data using an unsupervised machine learning algorithm;receiving, by the computer system, training data comprising a plurality of query representations, a plurality of search result ...

UNSUPERVISED LEARNING OF ENTITY REPRESENTATIONS USING GRAPHS

Techniques for implementing a learning semantic representations of sparse entities using unsupervised embeddings are disclosed herein. In some embodiments, a computer system accesses corresponding profile data of users indicating at least one entity of a first facet type associated with the user, and generating a graph data structure comprising nodes and edges based on the accessed profile data, with each node corresponding to a different entity indicated by the accessed profile data, and each edge directly connecting a different pair of nodes and indicating a number of users whose profile data indicates both entities of the pair of nodes. The computer system generating a corresponding embedding vector for the entities based on the graph data structure using an unsupervised machine learning algorithm. 1. A computer-implemented method comprising:for each one of a plurality of users of an online service, accessing, by a computer system having a memory and at least one hardware processor, corresponding profile data of the user stored on a database of the online service, the accessed profile data of each user indicating at least one entity of a first facet type associated with the user;generating, by the computer system, a graph data structure based on the accessed profile data, the generated graph data structure comprising a plurality of nodes and a plurality of edges, each one of the plurality of nodes corresponding to a different entity indicated by the accessed profile data, and each one of the plurality of edges directly connecting a different pair of the plurality of nodes and indicating a number of the plurality of users whose profile data indicates both entities of the pair of nodes that are directly connected by the edge;generating, by the computer system, a corresponding embedding vector for each one of the entities indicated by the accessed profile data based on the graph data structure using an unsupervised machine learning algorithm; andperforming, by the ...

USER NOTIFICATIONS BASED ON PROJECT CONTEXT

Disclosed herein are systems, methods, and non-transitory computer-readable media for providing user notifications based on a project context. The system may receive candidate attributes from candidate devices of a plurality of candidates and storing them in a candidate database, as well as user-entered attributes from a user device of a user. The system may then iteratively execute a number of operations that include performing a search for candidates in the candidate database by comparing project attributes with candidate attributes and providing user notification of newly-matched candidates that includes returning returned candidates that are matching candidates of the search results to the user based on the search. The system may generate context attributes from a context attribute source that include at least one of the user-entered attributes, candidate attributes of the returned candidates, candidate attributes of the selected candidate, candidate attributes of historical candidates, or project attributes of an additional project. 1. A method comprising , with a processor of a computer system:receiving, via a network interface, candidate attributes from candidate devices of a plurality of candidates and storing them in a candidate database in a. memory of the computer system;receiving, via the network interface, user-entered attributes from a user device of a user that are a first part of project attributes of a project and storing the project attributes in a project database in the memory; performing a search for candidates in the candidate database by comparing project attributes with candidate attributes and producing search results;', 'providing a user notification of newly-matched candidates that includes returning returned candidates that are matching candidates of the search results to the user based on the search, the returned candidates being candidates that the user has not previously seen, and marking the returned candidates as previously seen;', ' ...

Prefetching Using Dynamic User Model To Reduce Latency

A device includes a memory that stores a prefetching model. A control module receives a content page including one or more links each associated with selectable content and collects data associated with the content page. The collected data includes at least one of first data indicative of respective relationships between each of the links and a viewport of the device and second data indicative of characteristics of the viewport. The control module further assigns, using the prefetching model, respective scores to each of the links based on the collected data, and selectively generates, based on the assigned scores, a request to prefetch the selectable content associated with at least one of the links. 1. A device comprising:a memory that stores a prefetching model; and receives a content page including one or more links each associated with selectable content,', 'collects data associated with the content page, wherein the collected data includes at least one of (i) first data indicative of respective relationships between each of the links and a viewport of the device and (ii) second data indicative of characteristics of the viewport,', 'assigns, using the prefetching model, respective scores to each of the links based on the collected data, and', 'selectively generates, based on the assigned scores, a request to prefetch the selectable content associated with at least one of the links., 'a control module, in communication with the memory, that'}2. The device of claim 1 , wherein the links correspond to search results and the content page corresponds to a search engine result page.3. The device of claim 2 , wherein the collected data is indicative of user interaction with the viewport.4. The device of claim 3 , wherein the user interaction includes changing a position of the viewport relative to the search engine result page.5. The device of claim 2 , wherein the control module updates the collected data in response to a change in a position of the viewport relative ...

METHOD AND DEVICE FOR ON-CHIP REPETITIVE ADDRESSING

The present disclosure may include a method that comprises: partitioning data on an on-chip and/or an off-chip storage medium into different data blocks according to a pre-determined data partitioning principle, wherein data with a reuse distance less than a pre-determined distance threshold value is partitioned into the same data block; and a data indexing step for successively loading different data blocks to at least one on-chip processing unit according a pre-determined ordinal relation of a replacement policy, wherein the repeated data in a loaded data block being subjected to on-chip repetitive addressing. Data with a reuse distance less than a pre-determined distance threshold value is partitioned into the same data block, and the data partitioned into the same data block can be loaded on a chip once for storage, and is then used as many times as possible, so that the access is more efficient. 1. An on-chip repetitive addressing means , comprising:a data partitioning step for partitioning data on an on-chip storage medium and/or an off-chip storage medium into different data blocks according to a pre-determined data partitioning principle, wherein on the basis of the pre-determined data partitioning principle, the data with a reuse distance less than a pre-determined distance threshold value is partitioned into the same data block; anda data indexing step for successively loading the different data blocks to at least one on-chip processing unit according a pre-determined ordinal relation of a replacement policy, wherein the repeated data in a loaded data block being subjected to on-chip repetitive addressing.2. The on-chip repetitive addressing means according to claim 1 , wherein an index address for a data is consisted of a data block address and an in-block address;the data indexing step comprises successively loading different data blocks to the at least one on-chip processing unit according to the pre-determined ordinal relation of the replacement policy ...

ON-CHIP DATA PARTITIONING READ-WRITE METHOD, SYSTEM, AND DEVICE

The present invention is directed to the storage technical field and discloses an on-chip data partitioning read-write method, the method comprises: a data partitioning step for storing on-chip data in different areas, and storing the on-chip data in an on-chip storage medium and an off-chip storage medium respectively, based on a data partitioning strategy; a pre-operation step for performing an operational processing of an on-chip address index of the on-chip storage data in advance when implementing data splicing; and a data splicing step, for splicing the on-chip storage data and the off-chip input data to obtain a representation of the original data based on a data splicing strategy. Also provided are a corresponding on-chip data partitioning read-write system and device. Thus, read and write of repeated data can be efficiently realized, reducing memory access bandwidth requirements while providing good flexibility, thus reducing on-chip storage overhead. 1. An on-chip data partitioning read-write method , comprising:a data partitioning step for storing on-chip data in different areas, and storing the on-chip data in an on-chip storage medium and an off-chip storage medium respectively, based on a data partitioning strategy;a pre-operation step for performing an operational processing of an on-chip address index of the on-chip storage data in advance when implementing data splicing; anda data splicing step for splicing the on-chip storage data and the off-chip input data to obtain a representation of the original data based on a data splicing strategy.2. The on-chip data partitioning read-write method according to claim 1 , further comprising:a data storing step for storing and carrying the on-chip storage data of the on-chip storage medium and the off-chip input data from the off-chip storage medium;read-write ports are separated in the data storing step, and read and write of the data are independent from each other;the pre-operation step further comprises: ...

NEURAL NETWORK ACCELERATOR AND OPERATION METHOD THEREOF

A neural network accelerator and an operation method thereof applicable in the field of neural network algorithms are disclosed. The neural network accelerator comprises an on-chip storage medium for storing data externally transmitted or for storing data generated during computing; an on-chip address index module for mapping to a correct storage address on the basis of an input index when an operation is performed; a core computing module for performing a neural network operation; and a multi-ALU device for obtaining input data from the core computing module or the on-chip storage medium to perform a nonlinear operation which cannot be completed by the core computing module. By introducing a multi -ALU design into the neural network accelerator, an operation speed of the nonlinear operation is increased, such that the neural network accelerator is more efficient. 1. A neural network accelerator , comprising:an on-chip storage medium for storing data transmitted from an external of the neural network accelerator or for storing data generated during computation;an on-chip address index module for mapping to a correct storage address on the basis of an input index when an operation is performed;a core computing module for performing a linear operation of a neural network operation; anda multi-ALU device for obtaining input data from the core computing module or the on-chip storage medium to perform a nonlinear operation which cannot be performed by the core computing module.2. The neural network accelerator according to claim 1 , wherein the data generated during computation comprises a computation result or an intermediate computation result.3. The neural network accelerator according to claim 1 , wherein the multi-ALU device comprises: an input mapping unit claim 1 , a plurality of arithmetic logical units (ALUs) and an output mapping unit claim 1 ,the input mapping unit is configured for mapping the input data obtained from the on-chip storage medium or the core ...

STABILIZED PHOTO-ALIGNMENT LAYER FOR LIQUID CRYSTAL

A liquid-crystal photo-alignment layer that is composed of polymer stabilized azo dyes is provided, where a polymer network is introduced in the photo-alignment layer for stabilization. The photo-alignment layer is realized based on a two-step irradiation to first achieve molecule alignment to form the photo-alignment layer and then stabilize this layer. To realize the photo-alignment layer, a pre-determined surface of the substrate is first coated with a film of mixture. The mixture comprises an azo dye and a monomer preferably mixed in an optimal concentration of 0.67 wt/wt. The azo dye and the monomer have light-absorption peaks at different wavelengths such that photo-alignment of the azo-dye molecules and stabilization of the photo-alignment layer by polymerization of the monomer are achievable by two separate exposures of light to the film. 1. A liquid-crystal photo-alignment layer comprising a polymer-stabilized azo dye , the photo-alignment layer being formed by coating a substrate uniformly with an azo dye and a monomer and exposing to light for providing photo-alignment and stabilization such that good and stable photo-alignment for liquid crystals is achieved , wherein the photo-alignment and the stabilization have been established by two separate exposures of light.2. The liquid-crystal photo-alignment layer of claim 1 , wherein the polymer-stabilized azo dye alignment layer has been realized by mixing the azo dye and the monomer in an optimal concentration and then dissolving in a solvent.3. The liquid-crystal photo-alignment layer of claim 1 , wherein polymerization of the monomer has been done after defining a preferred orientation of an easy axis of the azo dye.4. The liquid-crystal photo-alignment layer of claim 1 , wherein polymerization of the monomer has been done by photo-cross-linking.5. The liquid-crystal photo-alignment layer of claim 1 , wherein polymerization of the monomer has been done by thermal imidization.6. The liquid-crystal photo- ...

Ion resonance excitation operation method and device by applying a quadrupolar electric field combined with a dipolar electric field

An ion resonance excitation operation method and device by applying a quadrupolar electric field combined with a dipolar electric field. The method includes applying a main RF to any pair of plates of the ion trap mass analyzer, and applying a quadrupolar excitation signal to any pair of plates, and applying a reverse phase dipolar excitation signal to any pair of plates. Also provided is an ion resonance excitation operation method and device by using a quadrupolar electric field combined with a dipolar electric field, which includes applying a positive main RF to a pair of electrode rods of the quadrupole, and applying a negative main RF to the other pair of electrode rods; applying a quadrupolar excitation signal to any pair of electrode rods, applying a reverse phase dipolar excitation signal to any pair of electrode rods.

Apparatus and methods for non-linear function operations

A nonlinear function operation device and method are provided. The device may include a table looking-up module and a linear fitting module. The table looking-up module may be configured to acquire a first address of a slope value k and a second address of an intercept value b based on a floating-point number. The linear fitting module may be configured to obtain a linear function expressed as y=k×x+b based on the slope value k and the intercept value b, and substitute the floating-point number into the linear function to calculate a function value of the linear function, wherein the calculated function value is determined as the function value of a nonlinear function corresponding to the floating-point number. 1. A nonlinear function operation device , comprising: 'acquire a first address of a slope value k and a second address of an intercept value b based on a floating-point number; and', 'a table looking-up module configured to'}a linear fitting module configured toobtain a linear function expressed as y=k×x+b based on the slope value k and the intercept value b, andsubstitute the floating-point number into the linear function to calculate a function value of the linear function, wherein the calculated function value is determined as the function value of a nonlinear function corresponding to the floating-point number.2. The device of claim 1 , further comprising:a slope and intercept storing module configured to store multiple slope values and multiple intercept values of a plurality of linear functions, wherein the plurality of linear functions are obtained through piecewise-linear fitting of the nonlinear function, wherein the table looking-up module comprises a selecting module configured to acquire the first address of the slope value k and the second address of the intercept value b based on the floating-point number.3. The device of claim 2 , wherein the table looking-up module is further configured to acquire the slope value k and the intercept value b ...

Apparatus for operations at maxout layer of neural networks

Aspects for maxout layer operations in neural network are described herein. The aspects may include a load/store unit configured to retrieve input data from a storage module. The input data may be formatted as a three-dimensional vector that includes one or more feature values stored in a feature dimension of the three-dimensional vector. The aspects may further include a pruning unit configured to divide the one or more feature values into one or more feature groups based on one or more data ranges and select a maximum feature value from each of the one or more feature groups. Further still, the pruning unit may be configured to delete, in each of the one or more feature groups, feature values other than the maximum feature value and update the input data with the one or more maximum feature values. 1. An apparatus for data pruning at a maxout layer of a neural network , comprising: 'wherein the input data is formatted as a three-dimensional vector that includes one or more feature values stored in a feature dimension of the three-dimensional vector; and', 'a load/store unit configured to retrieve input data from a storage module,'} divide the one or more feature values into one or more feature groups based on one or more data ranges,', 'select a maximum feature value from each of the one or more feature groups,', 'delete, in each of the one or more feature groups, feature values other than the maximum feature value, and', 'update the input data with the one or more maximum feature values., 'a pruning unit configured to'}2. The apparatus of claim 1 , wherein the input data further includes an abscissa and an ordinate.3. The apparatus of claim 1 , further comprising a data conversion unit configured to adjust a write sequence in storing the input data.4. The apparatus of claim 3 , wherein the load/store unit is further configured to store the one or more feature values prior to storing data in other dimensions of the input data in accordance with the adjusted write ...

CONTEXT AWARE DYNAMIC CANDIDATE POOL RETRIEVAL AND RANKING

Disclosed in some examples are methods, systems, and machine readable mediums which provide for retrieval, ranking, and display of candidates that are more likely to respond to employment inquiries in an employment search graphical user interface (GUI). The system may employ a machine learning algorithm which may calculate a score for each member of the social networking service that predicts, based upon one or more features how likely the individual is to respond to a message. In some examples, the candidates that are determined to be more likely to respond may be presented as a selectable option in the GUI. 1. A method for providing a graphical user interface (GUI) , the method comprising: ["for each particular member of a plurality of members of the social networking service, determining, using a machine learning algorithm, a response score for the particular member based upon one or more of: activity data describing the particular member's activity on the social networking service, and profile data, the response score indicative of a likelihood of the particular member to respond to an inquiry about a job from another member of the social networking service;", 'causing a search GUI to be displayed to a user, the search GUI including GUI elements that receive search criteria;', 'receiving through the search GUI, a search query with member search criteria;', 'determining a search context of the user;', 'determining a threshold based upon the search context;', 'responsive to the search query, identifying a set of members from the plurality of members based on the response score determined for each particular member from the set of members and the threshold; and', 'causing the set of members to be displayed to the user as search results for the search query., 'using one or more processors of a social networking service, executing operations of2. The method of claim 1 , wherein the search context comprises the search criteria.3. The method of claim 1 , wherein ...

PROCESSING APPARATUS AND PROCESSING METHOD

A processing device with dynamically configurable operation bit width, characterized by comprising: a memory for storing data, the data comprising data to be operated, intermediate operation result, final operation result, and data to be buffered in a neural network; a data width adjustment circuit for adjusting the width of the data to be operated, the intermediate operation result, the final operation result, and/or the data to be buffered; an operation circuit for operating the data to be operated, including performing operation on data to be operated of different bit widths by using an adder circuit and a multiplier; and a control circuit for controlling the memory, the data width adjustment circuit and the operation circuit. The device of the present disclosure can have the advantages of strong flexibility, high configurability, fast operation speed, low power consumption or the like. 1. A processing device with dynamically configurable operation bit width , comprising:a memory for storing data, the data comprising data to be operated, intermediate operation result, final operation result, and data to be buffered of a neural network;a data width adjustment circuit for adjusting the width of the data to be operated, the intermediate operation result, the final operation result, and/or the data to be buffered;an operation circuit for operating the data to be operated of the neural network; anda control circuit for controlling the memory, the data width adjustment circuit, and the operation circuit.2. The device according to claim 1 , wherein the operation circuit operating the data to be operated of the neural network comprises determining a type of a multiplier circuit and an adder circuit of the operation circuit according to the data to be operated so as to perform the operation.3. The device according to claim 1 , wherein the data width adjustment circuit comprises:an input data processing module, configured to adjust the data width of the data in the memory; ...

USE OF STATINS IN THE TREATMENT OF ISCHEMIC DISEASES

The invention discloses statins compounds used as local intraosseous application for the treatment of ischemic diseases, in particular, discloses local intraosseous application of statin compounds in the treatment of peripheral ischemic diseases or cardio-cerebrovascular ischemic diseases. 1. A composition comprising a statin or a pharmaceutically acceptable salt thereof , and a pharmaceutically acceptable carrier , wherein said composition is formulated for administration as a single dose administered locally into bone as a treatment for an ischemic disease , disorder or condition in a mammal.2. The composition according to claim 1 , wherein the statin is selected from the group consisting of simvastatin claim 1 , atorvastatin claim 1 , fluvastatin claim 1 , lovastatin claim 1 , pravastatin claim 1 , rosuvastatin claim 1 , pitavastatin claim 1 , bervastatin claim 1 , cerivastatin claim 1 , crilvastatin claim 1 , dalvastatin claim 1 , mevasatin claim 1 , tenivastatin and any combinations thereof.3. The composition of claim 2 , wherein the statin is simvastatin.4. The composition of claim 1 , wherein said composition further comprises a pharmaceutically acceptable adjuvant.5. The composition of claim 1 , wherein said pharmaceutically acceptable salt of statin is selected from the group consisting of hydrochloride claim 1 , hydrobromide claim 1 , hydriodate claim 1 , sulfate claim 1 , nitrate claim 1 , phosphate claim 1 , citrate claim 1 , mesylate claim 1 , trifluoroacetate claim 1 , and acetate.6. The composition of claim 1 , wherein the composition is selected from the group consisting of an injectable solution claim 1 , an injectable suspension liquid claim 1 , an injectable emulsion claim 1 , an injectable gel claim 1 , and an injectable implantable solid form.710-. (canceled)11. The composition of claim 1 , wherein the statin or pharmaceutically acceptable salt thereof is in a range from 0.1 mg to 50 mg.12. The composition of claim 1 , wherein the composition is ...

Method and device for soldering Hetero-Junction with intrinsic Thin-Film solar cells together to form string

A method for soldering hetero-junction with intrinsic thin layer solar cells together to form a string, includes soldering solar cells, and judging whether a temperature in a soldering chamber is within a preset temperature range every preset time; if the temperature is within the preset temperature range, continuing to solder; and if the temperature is beyond the preset temperature range, regulating the temperature in the soldering chamber to be within the preset temperature range, and continuing to solder the solar cells. 1. A method for soldering hetero junction with intrinsic thin-film solar cells together to form a string , comprising:soldering solar cells, and judging whether a temperature in a soldering chamber is within a preset temperature range every preset time;if the temperature is within the preset temperature range, continuing to solder; andif the temperature is beyond the preset temperature range, regulating the temperature in the soldering chamber to be within the preset temperature range, and continuing to solder the solar cells.2. The method according to claim 1 , before soldering solar cells claim 1 , further comprising:placing the solar cells on a transmission mechanism; andcontrolling the transmission mechanism to transmit the solar cells into the soldering chamber.3. The method for the hetero-junction with intrinsic thin-film solar cell according to claim 1 , after the soldering is completed claim 1 , further comprising:transmitting the solar cells out of the soldering chamber.4. The method for the hetero-junction with intrinsic thin-film solar cell according to claim 2 , wherein after placing the solar cells on the transmission mechanism claim 2 , and before controlling the transmission mechanism to transmit the solar cells into the soldering chamber claim 2 , the method further comprises:controlling the transmission mechanism to transmit the solar cells into a preheating region, so as to preheat a front side of the solar cells.5. The method ...

APPARATUS AND METHODS FOR FORWARD PROPAGATION IN FULLY CONNECTED LAYERS OF CONVOLUTIONAL NEURAL NETWORKS

Aspects for forward propagation in fully connected layers of a convolutional artificial neural network are described herein. The aspects may include multiple slave computation modules configured to parallelly calculate multiple groups of slave output values based on an input vector received via the interconnection unit. Further, the aspects may include a master computation module connected to the multiple slave computation modules via an interconnection unit, wherein the master computation module is configured to generate an output vector based on the intermediate result vector. 1. An apparatus for forward propagation in fully connected layers of a multilayer neural network , comprising:a master computation module configured to transmit an input vector via an interconnection unit; and to receive an intermediate result vector combined by the interconnection unit based on the multiple groups of slave output values calculated by the multiple slave computation modules, and', 'to generate an output vector based on the intermediate result vector., 'one or more slave computation modules configured to parallelly calculate multiple groups of slave output values based on the input vector received via the interconnection unit, wherein the master computation module is configured'}2. The apparatus of claim 1 , wherein the master computation module is configured to perform one operation selected from the group consisting of:adding a bias value to the intermediate result vector;activating the intermediate result vector with an activation function;outputting a predetermined value based on a comparison between the intermediate result vector and a random number; andpooling the intermediate result vector.3. The apparatus of claim 1 , wherein each of the slave computation modules includes a slave neuron caching unit configured to store the input vector.4. The apparatus of claim 1 , wherein the interconnection unit is structured as a binary tree including one or more levels claim 1 , ...

Device and Method for Performing Self-Learning Operations of an Artificial Neural Network

Aspects for self-learning operations of an artificial neural network are described herein. The aspects may include a master computation module configured to transmit an input vector via an interconnection unit and one or more slave computation modules connected to the master computation module via the interconnection unit. Each of the one or more slave computation modules may be configured to respectively store a column weight vector of a weight matrix and multiply the input vector with the column weight vector to generate a first multiplication result. The interconnection unit may be configured to combine the one or more first multiplication results into a first multiplication vector and transmit the first multiplication vector to the master computation module. 1. An apparatus for neural network operations , comprisinga master computation module configured to transmit an input vector via an interconnection unit; andone or more slave computation modules connected to the master computation module via the interconnection unit, respectively store a column weight vector of a weight matrix, and', 'multiply the input vector with the column weight vector to generate a first multiplication result, and, 'wherein each of the one or more slave computation modules is configured to combine the one or more first multiplication results into a first multiplication vector, and', 'transmit the first multiplication vector to the master computation module., 'wherein the interconnection unit is configured to2. The apparatus of claim 1 , wherein the master computation module is further configured to:add a first bias vector to the first multiplication vector to generate a first biased vector,activate the first biased vector by applying a first activation function to the first biased vector to generate a first activated vector, andsample the first activated vector by a Gibbs sampler to generate a first phase hidden layer vector.3. The apparatus of claim 2 ,wherein the master computation ...

Apparatus and methods for training in convolutional neural networks

Aspects for backpropagation of a convolutional neural network are described herein. The aspects may include a direct memory access unit configured to receive input data from a storage device and a master computation module configured to select one or more portions of the input data based on a predetermined convolution window. Further, the aspects may include one or more slave computation modules respectively configured to convolute one of the one or more portions of the input data with one of one or more previously calculated first data gradients to generate a kernel gradient, wherein the master computation module is further configured to update a prestored convolution kernel based on the kernel gradient.

METHOD AND ELECTRONIC APPARATUS FOR IMPLEMENTING MULTI-OPERATING SYSTEM

A method for implementing multi-operating system, applied to an electronic apparatus in which a Solid State Disk, SSD, is provided, the SSD including a plurality of partitions each of which corresponding to a unique logical snapshot table, and a plurality of operating systems being installed in different partitions respectively, wherein the method includes: determining a logical snapshot table corresponding to an operating system to be loaded currently as a first logical snapshot table during a Power On Self Test process of a basic input/output system; and determining a position of a partition in the SSD which corresponds to a reading/writing operation based on the first logical snapshot table if the reading/writing operation is performed on the SSD in a manner of Logical Block Addressing. 1. A method for implementing multi-operating system , applied to an electronic apparatus in which a Solid State Disk , SSD , is provided , the SSD comprising a plurality of partitions , and each partition corresponding to a unique logical snapshot table , and a plurality of operating systems being provided in the electronic apparatus and the plurality of operating systems being installed in different partitions , wherein the method comprises:determining a first logical snapshot table corresponding to an operating system to be loaded currently among a plurality of logical snapshot tables during a Power On Self Test process of a basic input/output system; anddetermining a position of a partition in the SSD which corresponds to a reading/writing operation according to the first logical snapshot table if the reading/writing operation is performed on the SSD in a manner of Logical Block Addressing.2. The method according to claim 1 , wherein determining a first logical snapshot table corresponding to an operating system to be loaded currently among a plurality of logical snapshot tables comprises:detecting a first operating system identifier corresponding to the operating system to be ...

APPARATUS AND METHODS FOR FORWARD PROPAGATION IN CONVOLUTIONAL NEURAL NETWORKS

Aspects for forward propagation of a convolutional artificial neural network are described herein. The aspects may include a direct memory access unit configured to receive input data from a storage device and a master computation module configured to select one or more portions of the input data based on a predetermined convolution window. Further, the aspects may include one or more slave computation modules respectively configured to convolute a convolution kernel with one of the one or more portions of the input data to generate a slave output value. Further still, the aspects may include an interconnection unit configured to combine the one or more slave output values into one or more intermediate result vectors, wherein the master computation module is further configured to merge the one or more intermediate result vectors into a merged intermediate vector. 1. An apparatus for forward propagation of a convolutional neural network , comprising: receive input data, and', 'select, in response to an instruction, one or more portions of the input data based on a predetermined convolution window;, 'a master computation module configured to'}one or more slave computation modules respectively configured to convolute a portion of a convolution kernel with one of the one or more portions of the input data to generate a slave output value; andan interconnection unit configured to combine the one or more slave output values into one or more intermediate result vectors, wherein the master computation module is further configured to merge the one or more intermediate result vectors into a merged intermediate vector.2. The apparatus of claim 1 , wherein each of the one or more slave computation modules includes a slave neuron caching unit configured to store one of the one or more portions of the input data.3. The apparatus of claim 1 , wherein each of the one or more slave computation modules includes a weight value caching unit configured to store the portion of the ...

Apparatus and Methods for Neural Network Operations Supporting Floating Point Numbers of Short Bit Length

Aspects for neural network operations with floating-point number of short bit length are described herein. The aspects may include a neural network processor configured to process one or more floating-point numbers to generate one or more process results. Further, the aspects may include a floating-point number converter configured to convert the one or more process results in accordance with at least one format of shortened floating-point numbers. The floating-point number converter may include a pruning processor configured to adjust a length of a mantissa field of the process results and an exponent modifier configured to adjust a length of an exponent field of the process results in accordance with the at least one format. 1. An apparatus for neural network operations , comprising:a neural network processor configured to process one or more floating-point numbers to generate one or more process results; and a pruning processor configured to adjust a length of a mantissa field of the process results, and', 'an exponent modifier configured to adjust a length of an exponent field of the process results in accordance with the at least one format., 'a floating-point number converter configured to convert the one or more process results in accordance with at least one format of shortened floating-point numbers, wherein the floating-point number converter includes'}2. The apparatus of claim 1 , further comprising a floating-point number analyzing processor configured to determine the at least one format of the shortened floating-point numbers claim 1 , wherein the floating-point number analyzing processor includes:a data extractor configured to collect one or more categories of the floating-point numbers;a data analyzer configured to statistically analyze the one or more categories of the floating-point numbers to determine a data range for each of the one or more categories of the floating-point numbers and a distribution pattern for each of the one or more categories ...

Apparatus and methods for backward propagation in neural networks supporting discrete data

Aspects for backpropagation of a multilayer neural network (MNN) in a neural network processor are described herein. The aspects may include a computation module configured to receive one or more groups of MNN data. The computation module may further include a master computation module configured to calculate an input gradient vector based on a first output gradient vector from an adjacent layer and based on a data type of each of the one or more groups of MNN data. Further still, the computation module may include one or more slave computation modules configured to parallelly calculate portions of a second output vector based on the input gradient vector calculated by the master computation module and based on the data type of each of the one or more groups of MNN data. 1. An apparatus for backpropagation of a multilayer neural network (MNN) , comprising: wherein the one or more groups of MNN data include input data and one or more weight values,', 'wherein at least a portion of the input data and the weight values are presented as discrete values, and', a master computation module configured to calculate an input gradient vector based on a first output gradient vector from an adjacent layer and based on a data type of each of the one or more groups of MNN data, and', 'one or more slave computation modules configured to parallelly calculate portions of a second output vector based on the input gradient vector calculated by the master computation module and based on the data type of each of the one or more groups of MNN data; and, 'wherein the computation module includes], 'a computation module configured to receive one or more groups of MNN data,'}a controller unit configured to decode an instruction that initiates a backpropagation process and transmit the decoded instruction to the computation module.2. The apparatus of claim 1 , wherein the interconnection unit is configured to combine the portions of the second output gradient vector to generate the second ...

Apparatus and method for executing recurrent neural network and lstm computations

Aspects for Long Short-Term Memory (LSTM) blocks in a recurrent neural network (RNN) are described herein. As an example, the aspects may include one or more slave computation modules, an interconnection unit, and a master computation module collectively configured to calculate an activated input gate value, an activated forget gate value, a current cell status of the current computation period, an activated output gate value, and a forward pass result.

METHOD AND SYSTEM FOR PROCESSING NEURAL NETWORK

The present disclosure provides a neural network processing system that comprises a multi-core processing module composed of a plurality of core processing modules and for executing vector multiplication and addition operations in a neural network operation, an on-chip storage medium, an on-chip address index module, and an ALU module for executing a non-linear operation not completable by the multi-core processing module according to input data acquired from the multi-core processing module or the on-chip storage medium, wherein the plurality of core processing modules share an on-chip storage medium and an ALU module, or the plurality of core processing modules have an independent on-chip storage medium and an ALU module. The present disclosure improves an operating speed of the neural network processing system, such that performance of the neural network processing system is higher and more efficient. 1. A system for processing a neural network , comprising:at least one on-chip storage medium for storing data transmitted from outside of a neural network processing system, or storing data generated during processing;at least one on-chip address index module for executing mapping according to an input index to a correct storage address during operation;a multi-core processing module composed of a plurality of core processing modules and for executing vector multiplication and addition operations in a neural network operation, andat least one ALU module for executing a non-linear operation not completable by the multi-core processing module according to input data acquired from the multi-core processing module or the on-chip storage medium,wherein the plurality of core processing modules share the on-chip storage medium and the ALU module, or the plurality of core processing modules have an independent on-chip storage medium and an ALU module.2. The processing system according to claim 1 , wherein the data generated during processing comprises a processing result or ...

File Reading Method, Storage Device And Electronic Device

A file reading method, storage device and electronic device are described. The file reading method is applied to an electronic device that includes a nonvolatile storage device as an internal storage device. The method includes determining a specific file in the electronic device as a hotspot file according to a predetermined condition; copying the determined hotspot file to the non-volatile storage device; and directly addressing the non-volatile storage device and reading the hotspot file from the nonvolatile storage device when receiving a request for reading the hotspot file. 1. A file reading method applied to an electronic device , the electronic device including a nonvolatile storage device as an internal storage device , the method comprising:determining a specific file in the electronic device as a hotspot file according to a predetermined condition;copying the determined hotspot file to the non-volatile storage device; anddirectly addressing the non-volatile storage device and reading the hotspot file from the nonvolatile storage device when receiving a request for reading the hotspot file.2. The file reading method according to claim 1 , wherein the receiving a request for reading the hotspot file further comprises making the page cache pointer of the node directly point to the nonvolatile storage device address space.3. The file reading method according to claim 1 , wherein determining a specific file in the electronic device as a hotspot file according to a predetermined condition comprises determining whether the file is a hotspot file according to the file accessing frequency claim 1 , wherein claim 1 , the file is determined to be a hotspot file when the file accessing frequency is high.4. The file reading method according to claim 1 , wherein determining a specific file in the electronic device as a hotspot file according to a predetermined condition comprises determining a hotspot file according to the file type claim 1 , wherein claim 1 , the file ...

PROCESSING APPARATUS AND PROCESSING METHOD

The present disclosure relates to a processing device including a memory configured to store data to be computed; a computational circuit configured to compute the data to be computed, which includes performing acceleration computations on the data to be computed by using an adder circuit and a multiplier circuit; and a control circuit configured to control the memory and the computational circuit, which includes performing acceleration computations according to the data to be computed. The present disclosure may have high flexibility, good configurability, fast computational speed, low power consumption, and other features. 1. A processing device comprising:a memory configured to store data, wherein the data include data to be computed in a neural network;a computational circuit configured to compute the data to be computed in the neural network, including performing computations on the data to be computed in the neural network with different computation bit widths by using an adder circuit and a multiplier circuit; anda control circuit configured to control the memory and the computational circuit, including determining a type of the multiplier circuit and a type of the adder circuit of the computational circuit according to the data to be computed so as to perform computations.2. The device of claim 1 , wherein the memory includes:an input storage module configured to store the data to be computed in the neural network,an output storage module configured to store a computation result,a neuron storage module configured to store neuron parameters,a synaptic storage module configured to store synaptic parameters, and wherein the output storage module further includes:', 'an intermediate computation result storage sub-module configured to store an intermediate computation result, and', 'a final computation result storage sub-module configured to store a final computation result., 'a caching module configured to cache data,'}3. The device of claim 2 , wherein the ...

PROCESSING APPARATUS AND PROCESSING METHOD

The present disclosure relates to a fused vector multiplier for computing an inner product between vectors, where vectors to be computed are a multiplier number vector {right arrow over (A)}{AN . . . A2A1A0} and a multiplicand number {right arrow over (B)} {BN . . . B2B1B0}, {right arrow over (A)} and {right arrow over (B)} have the same dimension which is N+1. The multiplier includes: N+1 multiplication sub-units configured to perform multiplication on each dimension of a vector respectively, and take lower n bits of the multiplier number vector for multiplication each time, where the n bits are removed from the binary number of each dimension of the multiplier number vector after the n bits are taken, and n is larger than 1 and less than N+1; an adder tree configured to perform addition on results of N+1 multiplication sub-units obtained from a same operation each time; and a result register configured to hold a result of every addition performed by the adder tree and send the result to the adder tree for next computation. 1. A fused vector multiplier for computing an inner product between vectors , comprising: wherein the multiplier number vector includes one or more first elements and the multiplicand vector includes one or more second elements, and', respectively retrieve one or more least significant bits of each of the one or more first elements,', 'multiply each of the one or more least significant bits with a respective one of the one or more second elements, and', 'remove the retrieved one or more least significant bits from each of the one or more first elements;, 'wherein each of the multiplication sub-units is configured to], 'multiple multiplication sub-units configured to perform multiplication between a multiplier number vector and a multiplicand vector;'}an adder tree configured to add results output from the multiplication sub-units; anda result register configured to store addition results output from the adder tree and feed the addition results ...

Conversion Method, Device, Computer Equipment, and Storage Medium

A model conversion method is disclosed. The model conversion method includes obtaining model attribute information of an initial offline model and hardware attribute information of a computer equipment, determining whether the model attribute information of the initial offline model matches the hardware attribute information of the computer equipment according to the initial offline model and the hardware attribute information of the computer equipment and in the case when the model attribute information of the initial offline model does not match the hardware attribute information of the computer equipment, converting the initial offline model to a target offline model that matches the hardware attribute information of the computer equipment according to the hardware attribute information of the computer equipment and a preset model conversion rule. 1. A model conversion method for converting an initial offline model to a target offline model matching a plurality of hardware attributes of a computer system , the method comprising:obtaining, by an I/O interface of the computer system, the initial offline model and storing the initial offline model to a local memory of the computer system;obtaining, by a processor of the computer system, the plurality of hardware attributes of the computer system;determining, by the processor of the computer system, whether model attributes of the initial offline model match the plurality of hardware attributes of the computer system according to the initial offline model and the plurality of hardware attributes of the computer system; andin the case when the model attributes of the initial offline model do not match the plurality of hardware attributes of the computer system, converting, by the processor, the initial offline model to a target offline model that matches the plurality of hardware attributes of the computer system according to the plurality of hardware attributes of the computer system and a preset model conversion ...

TASK PARALLEL PROCESSING METHOD, APPARATUS AND SYSTEM, STORAGE MEDIUM AND COMPUTER DEVICE

Computer systems, data processing methods, and computer-readable media are provided to run original networks. An exemplary computer system includes first and second processors and first and second memories. The first memory stores offline models and corresponding input data of a plurality of original networks, and a runtime system configured to run on the first processor. The second memory stores an operating system configured to run on the first processor or the second processor. When the runtime system runs on the first processor, the runtime system obtains an offline model and corresponding input data of an original network from the first memory and controls the second processor to run the offline model of the original network. The offline model of the original network includes model parameters, instructions, and interface data of respective computation nodes of the original network. 1. A computer system , comprising:first and second processors; andfirst and second memories; the first memory stores offline models and corresponding input data of a plurality of original networks, and a runtime system configured to run on the first processor;', 'the second memory stores an operating system configured to run on the first processor or the second processor;', 'the runtime system is a secure runtime system built based on a trusted operating environment;', 'the first memory is a secure storage medium;', 'when the runtime system runs on the first processor, the runtime system obtains an offline model and corresponding input data of an original network from the first memory, and controls the second processor to run the offline model of the original network; and', 'the offline model of the original network comprises model parameters, instructions, and interface data of respective computation nodes of the original network., 'wherein2. The computer system of claim 1 , wherein the runtime system claim 1 , when runs on the first processor claim 1 , causes the first processor to ...

SCHEDULING IN JOB EXECUTION

The present invention relates to a method, apparatus, and computer program product for scheduling in job execution. According to embodiments of the present invention, there is provided a method for scheduling a plurality of job slots shared by one or more pre-processors and one or more post-processors in job execution, wherein the data generated by the pre-processor(s) will be fed to the post-processor(s) for processing. The method comprises: determining an overall data generation speed of the pre-processor(s); determining an overall data consumption speed of the post-processor(s); and scheduling allocation of at least one of the job slots between the pre-processor(s) and the post-processor(s) based on the overall data generation speed and the overall data consumption speed. Corresponding apparatus is disclosed as well. 1. A method for scheduling a plurality of job slots shared by one or more pre-processors and one or more post-processors in job execution , wherein data generated by the one or more pre-processors will be fed to the one or more post-processors for processing , the method comprising:determining, by one or more computer processors, an overall data generation speed of the one or more pre-processors;determining, by one or more computer processors, an overall data consumption speed of the one or more post-processors; andscheduling, by one or more computer processors, allocation of at least one of the plurality of job slots between the one or more pre-processors and the one or more post-processors based, at least in part, on the overall data generation speed and the overall data consumption speed, to lower an imbalance between the overall data generation speed and the overall data consumption speed.2. The method according to claim 1 , wherein scheduling allocation of at least one of the plurality of job slots between the one or more pre-processors and the one or more post-processors based claim 1 , at least in part claim 1 , on the overall data generation ...

STABLE FLIGHT CONTROL METHOD FOR MULTI-ROTOR UNMANNED AERIAL VEHICLE BASED ON FINITE-TIME NEURODYNAMICS

Provided is a stable flight control method for a multi-rotor unmanned aerial vehicle based on finite-time neurodynamics, comprising the following implementation process: 1) acquiring real-time flight orientation and attitude data through airborne sensors, and analyzing and processing kinematic problems of the aerial vehicle through an airborne processor to establish a dynamics model of the aerial vehicle; 2) designing a finite-time varying-parameter convergence differential neural network solver according to a finite-time varying-parameter convergence differential neurodynamics design method; 3) solving output control parameters of motors of the aerial vehicle through the finite-time varying-parameter convergence differential neural network solver using the acquired real-time orientation and attitude data; and 4) transmitting results to speed regulators of the motors of the aerial vehicle to control the motion of the unmanned aerial vehicle. Based on the finite-time varying-parameter convergence differential neurodynamics method, the invention can approximate the correct solution of the problem in a quick, accurate and real-time way, and can well solve a variety of time-varying problems such as matrix, vector, algebra and optimization. 1. A stable flight control method for a multi-rotor unmanned aerial vehicle based on finite-time neurodynamics , the control method comprising the steps of:{'b': '1', 'S, acquiring real-time flight orientation and attitude data of the multi-rotor unmanned aerial vehicle through sensors thereof, and analyzing and processing kinematic problems of the aerial vehicle correspondingly through an airborne processor to establish a dynamics model of the aerial vehicle;'}{'b': '2', 'S, designing a finite-time varying-parameter convergence differential neural network solver for the dynamics model of the multi-rotor aerial vehicle according to a finite-time varying-parameter convergence differential neurodynamics design method;'}{'b': '3', 'S, ...

APPARATUS AND METHODS FOR FORWARD PROPAGATION IN CONVOLUTIONAL NEURAL NETWORKS

Aspects for forward propagation of a convolutional artificial neural network are described herein. The aspects may include a direct memory access unit configured to receive input data from a storage device and a master computation module configured to select one or more portions of the input data based on a predetermined convolution window. Further, the aspects may include one or more slave computation modules respectively configured to convolute a convolution kernel with one of the one or more portions of the input data to generate a slave output value. Further still, the aspects may include an interconnection unit configured to combine the one or more slave output values into one or more intermediate result vectors, wherein the master computation module is further configured to merge the one or more intermediate result vectors into a merged intermediate vector. 1. An apparatus for neural network operations , comprising:a controller circuit configured to receive an instruction; and receive input data,', 'select, in response to an instruction, one or more portions of the input data based on a predetermined convolution window, wherein the instruction includes a first address of the one or more portions of the input data, a first size of the one or more portions of the input data, a second address of a portion of a convolution kernel, and a second size of the portion of the convolution kernel,', 'convolute the portion of the convolution kernel with one of the one or more portions of the input data to generate a slave output value,', 'combine the one or more slave output values into one or more intermediate result vectors, and', 'merge the one or more intermediate result vectors into a merged intermediate vector., 'a computation circuit configured to2. The apparatus of claim 1 , wherein the computation circuit includes a master computation circuit configured to:receive the input data,select, in response to the instruction, the one or more portions of the input data ...

APPARATUS AND METHODS FOR TRAINING IN CONVOLUTIONAL NEURAL NETWORKS

Aspects for backpropagation of a convolutional neural network are described herein. The aspects may include a direct memory access unit configured to receive input data from a storage device and a master computation module configured to select one or more portions of the input data based on a predetermined convolution window. Further, the aspects may include one or more slave computation modules respectively configured to convolute one of the one or more portions of the input data with one of one or more previously calculated first data gradients to generate a kernel gradient, wherein the master computation module is further configured to update a prestored convolution kernel based on the kernel gradient. 1. An apparatus for backpropagation of a convolutional neural network , comprising:a controller unit configured to receive an instruction; and receive input data,', 'select one or more portions of the input data based on a predetermined convolution window in response to an instruction,', 'respectively convolute one of the one or more portions of the input data with one of one or more calculated first data gradients to generate a kernel gradient, and', 'update a prestored convolution kernel based on the kernel gradient., 'a computation circuit configured to2. The apparatus of claim 1 , wherein the computation circuit includes a master computation module configured to:receive the input data,select the one or more portions of the input data based on the predetermined convolution window in response to the instruction, andupdate a prestored convolution kernel based on the kernel gradient.3. The apparatus of claim 1 , wherein the computation circuit include one or more slave computation modules respectively configured to convolute the one of the one or more portions of the input data with the one of one or more calculated first data gradients to generate the kernel gradient.4. The apparatus of claim 3 , wherein the one or more slave computation modules are respectively ...

COMPILATION OF APPLICATION INTO MULTIPLE INSTRUCTION SETS FOR A HETEROGENEOUS PROCESSOR

Techniques generally described are related to a method to compile code for a heterogeneous multi-core processor that includes a first core and a second core. The method may include receiving, by a multi-core compilation system, a set of source code that includes a plurality of code segments, wherein the multi-core compilation system is configured to compile the set of source code and generate an executable program that is executable by the heterogeneous multi-core processor. The method may include generating, by the multi-core compilation system, a first instruction set based on a specific code segment selected from the plurality of code segments, wherein the first instruction set is executable by the first core of the heterogeneous multi-core processor. The method may further include, in response to a determination that a performance indicator associated with the first core executing the first instruction set is above a particular threshold, generating, by the multi-core compilation system, a second instruction set based on the specific code segment, wherein the second instruction set is executable by the second core of the heterogeneous multi-core processor, and the first instruction set and the second instruction set are implemented in the executable program. 1. A method to compile code for a heterogeneous multi-core processor that includes a first core and a second core , the method comprising:receiving, by a multi-core compilation system, a set of source code that includes a plurality of code segments, wherein the multi-core compilation system is configured to compile the set of source code and generate an executable program that is executable by the heterogeneous multi-core processor;generating, by the multi-core compilation system, a first instruction set based on a specific code segment selected from the plurality of code segments, wherein the first instruction set is executable by the first core of the heterogeneous multi-core processor; andin response to a ...

Apparatus and Methods for Neural Network Operations Supporting Fixed Point Numbers of Short Bit Length

Aspects for neural network operations with fixed-point number of short bit length are described herein. The aspects may include a fixed-point number converter configured to convert one or more first floating-point numbers to one or more first fixed-point numbers in accordance with at least one format. Further, the aspects may include a neural network processor configured to process the first fixed-point numbers to generate one or more process results.

FACTORED MODEL FOR SEARCH RESULTS AND COMMUNICATIONS BASED ON SEARCH RESULTS

In an example embodiment, two machine learned models are trained. One is trained to output a probability that a searcher having a member profile in a social networking service will select a potential search result. The other is trained to output a probability that a member corresponding to a potential search result will respond to a communication from a searcher. Features may be extracted from a query, information about the searcher, and information about the member corresponding to the potential search result and fed to the machine learned models. The outputs of the machine learned models can be combined and used to rank search results for returning to the searcher. 1. A system for returning search results in an online computer system , the system comprising:a computer readable medium having instructions stored there on, which, when executed by a processor, cause the system to:obtain a plurality of candidate search results in response to a query corresponding to a searcher;obtain a member profile, in a social networking service, corresponding to the searcher; obtain activity and usage information for a member corresponding to the candidate search result;', 'parse the candidate search result, the activity and usage information for the member corresponding to the candidate search result, and the member profile corresponding to the searcher to extract a first set of one or more features and a second set of one or more features;', 'feed the extracted first set of one or more features into a first machine learned model trained to output a probability that a searcher having the sample member profile will select a potential search result;', 'feed the extracted second set of one or more features into a second machine learned model trained to output a probability that first member corresponding candidate search result will respond to a communication from a searcher;', 'combine outputs of the first machine learned model and the second machine learned model to form a combined ...

Apparatus and methods for backward propagation in neural networks supporting discrete data

Aspects for backpropagation of a multilayer neural network (MNN) in a neural network processor are described herein. The aspects may include a computation module configured to receive one or more groups of MNN data. The computation module may further include a master computation module configured to calculate an input gradient vector based on a first output gradient vector from an adjacent layer and based on a data type of each of the one or more groups of MNN data. Further still, the computation module may include one or more slave computation modules configured to parallelly calculate portions of a second output vector based on the input gradient vector calculated by the master computation module and based on the data type of each of the one or more groups of MNN data. 1. An apparatus for backpropagation of a multilayer neural network (MNN) , comprising: wherein the one or more groups of MNN data include input data and one or more weight values,', 'wherein at least a portion of the input data and the weight values are presented as discrete values, and', a master computation module configured to calculate an input gradient vector based on a first output gradient vector from an adjacent layer and based on a data type of each of the one or more groups of MNN data, and', 'one or more slave computation modules configured to parallelly calculate portions of a second output vector based on the input gradient vector calculated by the master computation module and based on the data type of each of the one or more groups of MNN data; and, 'wherein the computation module includes], 'a computation module configured to receive one or more groups of MNN data,'}a controller unit configured to decode an instruction that initiates a backpropagation process and transmit the decoded instruction to the computation module.2. The apparatus of claim 1 , wherein the interconnection unit is configured to combine the portions of the second output gradient vector to generate the second ...

RAYCAST CALIBRATION FOR ARTIFICIAL REALITY HEAD-MOUNTED DISPLAYS

Raycast-based calibration techniques are described for determining calibration parameters associated with components of a head mounted display (HMD) of an augmented reality (AR) system having one or more off-axis reflective combiners. In an example, a system comprises an image capture device and a processor executing a calibration engine. The calibration engine is configured to determine correspondences between target points and camera pixels based on images of the target acquired through an optical system, the optical system including optical surfaces and an optical combiner. Each optical surface is defined by a difference of optical index on opposing sides of the surface. At least one calibration parameter for the optical system is determined by mapping rays from each camera pixel to each target point via raytracing through the optical system, the raytracing being based on the index differences, shapes, and positions of the optical surfaces relative to the one or more cameras. 1. A method of calibrating an optical system comprising:determining a plurality of correspondences between a plurality of target points of a target and a plurality of camera pixels of one or more cameras based on one or more images of the target acquired by the camera through an optical system, the optical system including a plurality of optical surfaces and an optical combiner, wherein each of the plurality of optical surfaces is defined by a difference of optical index on opposing sides of the surface; anddetermining at least one calibration parameter for the optical system by mapping a plurality of rays from each of the plurality of camera pixels to each of the plurality of target points via raytracing through the optical system, wherein the raytracing is based on the index differences of the plurality of optical surfaces, the shapes of the plurality of optical surfaces, and the positions of the plurality of optical surfaces relative to the one or more cameras.2. The method of claim 1 , ...

Selective Nucleic Acid Fragment Recovery

The invention provides methods and kits for nucleic acid purification and fragment selection and recovery. By adjusting salt concentrations and/or pH of the binding buffers, only nucleic acid fragments with desired size ranges are able to reversibly and non-specifically bind to a solid surface in certain salt concentrations and/or pH conditions, and can be subsequently eluted and/or recovered from the solid surface in water and/or a low salt eluting buffer. 1. A method to selectively recover nucleic acid fragments with a desired size range comprising:a) mixing target nucleic acids with a binding buffer having a selected pH and salt concentration to allow the nucleic acid fragments with the desired size range to be bound to a solid surface;b) applying the mixture from step a) to the solid surface so that the nucleic acid fragments with the desired size range reversibly and non-specifically bind to the solid surface; andc) selectively recovering the nucleic acid fragments with the desired size range by eluting the bound nucleic acid fragments from the solid surface with an eluting buffer.2. The method of claim 1 , wherein said binding buffer comprises a chaotropic salt selected from the group consisting of guanidium hydrochloride (GHCl) in a concentration between 0.8-5.0 M claim 1 , guanidium thiocyanate (GITC) in a concentration between 0.8-5.0 M claim 1 , sodium iodide (NaI) in a concentration between 0.8-7.0 M claim 1 , sodium perchlorate in a concentration between 1.0-7.0 M claim 1 , and mixtures thereof.3. The method of claim 2 , wherein said chaotropic salt concentration of said binding buffer is adjusted by water in an amount that allows said nucleic acid fragments with the desired size range being bound to and subsequently eluted from said solid surface.4. The method of claim 3 , wherein said solid surface is silica membrane filter column or silica coated magnetic microparticles.5. The method of claim 4 , wherein said nucleic acid fragments are DNA fragments ...

LEAKAGE PROTECTIVE PLUG

A leakage protective plug and a tripping mechanism thereof, comprising a housing, and a tripping mechanism and a conducting structure arranged within the housing, wherein the tripping mechanism comprises a restorable reset button, a tripping bracket, and a tripping coil and a pressing arm arranged in the tripping bracket, wherein the reset button is sleeved in the pressing arm in which a snap fitting mechanism is positioned such that the pressing arm and the reset button may be snapped together, and the pressing arm and the reset button may be disengaged under the control of the tripping coil. It has a simple and reliable structure, and avoids technical problems including poor contact of supply circuit and failure of power leakage protecting functions due to un-complete tripping which are resulted from the deformation of the tripping plate caused by the movement of the tripping iron core and by other external forces. 1. A leakage protective plug , comprising a housing , and a tripping mechanism and a conducting structure which are arranged within the housing , wherein the tripping mechanism comprises a restorable reset button , a tripping bracket , and a tripping coil and a pressing arm which are arranged in the tripping bracket , wherein the reset button is sleeved in the pressing arm , a snap fitting mechanism is positioned in the pressing arm for snapping the pressing arm and the reset button together , and the pressing arm and the reset button are disengageable under the control of the tripping coil.2. The leakage protective plug according to claim 1 , characterized in that: the snap fitting mechanism of the pressing arm comprises a snap fitting sliding groove arranged in the pressing arm and a snap fitting part arranged in the snap fitting sliding groove claim 1 , and the reset button is upwardly and downwardly slidable in the snap fitting sliding groove and is snap fitted with the snap fitting part.3. The leakage protective plug according to claim 1 , ...

APPARATUS AND METHODS FOR NEURAL NETWORK OPERATIONS SUPPORTING FIXED POINT NUMBERS OF SHORT BIT LENGTH

Aspects for neural network operations with fixed-point number of short bit length are described herein. The aspects may include a fixed-point number converter configured to convert one or more first floating-point numbers to one or more first fixed-point numbers in accordance with at least one format. Further, the aspects may include a neural network processor configured to process the first fixed-point numbers to generate one or more process results. 1. An apparatus for neural network operations , comprising:a fixed-point number converter configured to convert one or more first floating-point numbers to one or more first fixed-point numbers in accordance with at least one format;a neural network processor configured to process the first fixed-point numbers to generate one or more process results, anda floating-point number analyzing processor configured to determine the at least format of the fixed-point numbers by statistically analyzing one or more categories of the first floating-point numbers to determine a distribution pattern for each of the one or more categories over one or more data ranges.2. The apparatus of claim 1 , wherein the floating-point number analyzing processor includes:a data extractor configured to collect one or more categories of the first floating-point numbers; anda data analyzer configured to statistically analyze the one or more categories of the first floating-point numbers to determine the distribution pattern for each of the one or more categories over the one or more data ranges.3. The apparatus of claim 2 , wherein the floating-point number analyzing processor further includes a format determiner configured to determine the at least one format of fixed-point numbers respectively corresponding to each of the one or more categories.4. The apparatus of claim 3 , wherein the format determiner is further configured to determine a point location that indicates a location of a binary point for each of the at least one format of fixed-point ...

NONLINEAR MODELS FOR MEMBER SEARCHING

A system and method includes receiving a search query and obtaining, from a database, member data of a member. For each of a plurality of nonlinear models, the nonlinear model is traversed based on a comparison of characteristics to conditions to obtain a score, wherein, among the nonlinear models, at least one characteristic is an inferred characteristic based on at least one of: activities by the member in an online networking system; and connections by the member in the online networking system. The score obtained from each of the nonlinear models is combined to obtain a combined score and a user interface to displays information related to the member based, at least in part on the combined score. 1. A processor-implemented method , comprising:receiving a search query;obtaining, from a database of an online social networking system, member data of a member of the online social networking system;for each of a plurality of nonlinear models, traversing the nonlinear model based on a comparison of individual characteristics to an associated condition to obtain a score, wherein, among the nonlinear models, at least one characteristic is an inferred characteristic based on at least one of: activities by the member in the online social networking system; and connections by the member in the online social networking system;combining the scores obtained from each of the nonlinear models to obtain a combined score; andcausing a user interface to display information related to the member based, at least in part on the combined score.2. The method of claim 1 , wherein each of the nonlinear models includes a plurality of hierarchically related nodes claim 1 , the nodes including characteristic nodes and result nodes claim 1 , each characteristic node including a characteristic and a condition to traverse to another node claim 1 , each result node including a score claim 1 , wherein traversing the nonlinear model is by:comparing, for each characteristic node arrived at, the ...

OPERATION APPARATUS AND METHOD

The present disclosure provides an operation device, comprising: an operation module for executing a neural network operation; and a power conversion module connected to the operation module, for converting input neuron data and/or output neuron data of the neural network operation into power neuron data. The present disclosure further provides an operation method. The operation device and method according to the present disclosure reduce the cost of storage resources and computing resources and increase the operation speed. 1. An operation device , comprising:an operation module for executing a neural network operation; anda power conversion module connected to the operation module, for converting input neuron data and/or output neuron data of the neural network operation into power neuron data.2. The operation device according to claim 1 , wherein the power conversion module comprises:a first power conversion unit for converting the neuron data output from the operation module into the power neuron data; anda second power conversion unit for converting the neuron data input to the operation module into the power neuron data.3. The operation device according to claim 2 , wherein the operation module further comprises a third power conversion unit for converting the power neuron data into non-power neuron data.4. The operation device according to claim 1 , further comprising:a storage module for storing data and operational instructions; anda control module for controlling interactions of the data and the operational instructions, the control module being configured to receive the data and the operational instructions sent by the storage module and decode the operational instructions into operational microinstructions;wherein the operation module comprises an operation unit for receiving the data and the operational microinstructions sent by the control module, and performing the neural network operation on the received data according to the operational ...

Method and Device for Processing Random Access Preamble

The disclosure provides a method and device for processing a random access preamble. The method includes: determining a Guard Time (GT) required by communication between a User Equipment (UE) and a base station according to a cell coverage radius, wherein the cell coverage radius is larger than 100 kilometers; and determining the random access preamble transmitted by the UE to the base station according to the GT, wherein the random access preamble is transmitted on an ordinary sub-frame in a radio frame during the communication between the UE and the base station. By the method and the device, the problem that the cell coverage radius of a single station cannot exceed 100 kilometers in a related technology is solved, and effects of enlarging the cell coverage radius of the single station and meeting a requirement on an ultra-large cell coverage radius in a special scenario are further achieved.

Data Accessing Method And Data Accessing Apparatus

A data accessing method and a data accessing apparatus are provided. The method includes acquiring a size of a minimum access unit of a storage device; reading first data to be written into the storage device; determining whether a size of the first data is equal to an integer multiple of the size of the minimum access unit; and if the size of the first data is equal to an integer multiple of the size of the minimum access unit, storing the first data into the integer multiple of the minimum access unit in alignment; otherwise, adding padding data to the first data such that the size of the first data is equal to the integer multiple of the size of the minimum access unit, and storing the first data into the integer multiple of the minimum access unit in alignment. 1. A data accessing method , comprising:acquiring a size of a minimum access unit of a storage device;reading first data to be written into the storage device;determining whether a size of the first data is equal to an integer multiple of the size of the minimum access unit;if the size of the first data is equal to an integer multiple of the size of the minimum access unit, storing the first data into the integer multiple of the minimum access unit in alignment; andotherwise, adding padding data to the first data such that the size of the first data is equal to the integer multiple of the size of the minimum access unit, and storing the first data into the integer multiple of the minimum access unit in alignment.2. The data accessing method according to claim 1 , wherein the step of storing the first data into the integer multiple of the minimum access unit in alignment comprises:determining a first start address to start storing the first data;determining whether the first start address is a start address of one minimum access unit;if yes, storing the first data into the integer multiple of the minimum access unit in alignment by starting from the first start address; andotherwise, shifting the first ...

TLB DEVICE SUPPORTING MULTIPLE DATA STREAMS AND UPDATING METHOD FOR TLB MODULE

Aspects of managing Translation Lookaside Buffer (TLB) units are described herein. The aspects may include a memory management unit (MMU) that includes one or more TLB units and a control unit. The control unit may be configured to identify one from the one or more TLB units based on a stream identification (ID) included in a received virtual address and, further, to identify a frame number in the identified TLB unit. A physical address may be generated by the control unit based on the frame number and an offset included in the virtual address. 1. A memory management unit , comprising: wherein each of the TLB units corresponds to one of multiple data types,', 'wherein each of the TLB units includes a plurality of page numbers and a plurality of frame numbers that respectively correspond to the plurality of page numbers; and, 'one or more Translation Lookaside Buffer (TLB) units,'} receive a virtual address that includes a stream identification (ID), a virtual page number, and an offset, wherein the stream ID identifies one of the multiple data types,', 'identify one of the one or more TLB units that corresponds to the stream ID in the virtual address based on the data type indicated by the stream ID,', 'determine a TLB hit when one of the plurality of page numbers in the identified TLB unit matches the virtual page number in the virtual address,', 'based on the determination of the TLB hit, identify one of the plurality of page numbers that corresponds to the virtual page number in the virtual address,', 'identify the frame number that corresponds to the identified page number, and', 'generate a physical address that includes the frame number and the offset., 'a control unit configured to2. The memory management unit of claim 1 , wherein the control unit is further configured to:determine a TLB miss when none of the page numbers in the identified TLB unit matches the virtual page number in the virtual address,based on the determination of the TLB miss, identify an ...

BRAKE/CLUTCH DEVICE FOR MANUAL HOIST

A brake/clutch device for a manual hoist provides an externally-viewable indicator showing the wear status of friction elements within the hoist. The brake/clutch device has a brake block, a brake ratchet, a reversing ratchet, a torsion spring block, a hand wheel and a fastener mounted onto a main shaft of the manual hoist. Friction elements on the brake ratchet provide braking against sudden drops of a load. Ball and spring devices mounted in the hand wheel contact detents and grooves that gradually increase in depth in the torsion spring block depending on the relative positions of the hand wheel and the torsion spring block. The hand wheel has an indicator mark that may align with indicator areas of the torsion spring block to indicate an extent of wear of the friction elements 1. A braking/clutching device of a manual hoist , comprising:an assembly including a brake block, a brake ratchet, a reversing ratchet, a torsion spring block, a hand wheel and a fastener circumferentially disposed successively on a long shaft of the manual hoist, the brake block fixing one end of the assembly and the fastener fixing the other end thereof;the brake block and the long shaft being fixed circumferentially;a friction disk provided on both sides of the brake ratchet;the reversing ratchet and the long shaft being connected by threads;a bayonet connection connecting the torsion spring block and the reversing ratchet in a are fixed circumferential connection;a torsion spring mounted onto the long shaft at a position located between the torsion spring block and the hand wheel, the torsion spring having a first end coupled to the torsion spring block and the torsion spring having a second end coupled to the hand wheel;steel ball and spring device disposed between the torsion spring block and the hand wheel and mounted in the hand wheel, andthe torsion spring block defining concave detents on an end face of the torsion spring block;the hand wheel and the long shaft being fixed ...

OPERATION UNIT, METHOD AND DEVICE CAPABLE OF SUPPORTING OPERATION DATA OF DIFFERENT BIT WIDTHS

Aspects for supporting operation data of different bit widths in neural networks are described herein. The aspects may include a processing module that includes one or more processors. The processor may be capable of processing data of one or more respective bit-widths. Further, the aspects may include a determiner module configured to receive one or more instructions that include one or more operands and one or more width fields. The operands may correspond to one or more operand types and each of the width fields may indicate an operand bit-width of one operand type. The determiner module may be further configured to identify at least one operand bit-widths that is greater than each of the bit-widths. In addition, the aspects may include a processor combiner configured to designate a combination of two or more of the processors to process the operands. 1. An apparatus for neural network processing , comprising:a processing module that includes one or more processors, wherein the one or more processors are capable of processing data of one or more respective bit-widths; [ wherein the one or more operands correspond to one or more operand types, and', 'wherein each of the one or more width fields indicates an operand bit-width of one of the one or more operand types, and, 'receive one or more instructions that include one or more operands and one or more width fields,'}, 'identify at least one of the one or more operand bit-widths that is greater than each of the one or more bit-widths,', 'transmit the operands that correspond to the at least one of the one or more operand bit-widths; and, 'a determiner module configured toa processor combiner configured to designate a combination of two or more of the processors to process the operands that correspond to the at least one of the operand bit-widths.2. The apparatus of claim 1 , wherein the one or more operands and the one or more width fields are included in one of the one or more instructions.3. The apparatus of ...

SINGLE LOCAL OSCILLATOR IN A MULTI-BAND FREQUENCY DIVISION DUPLEX TRANSCEIVER

Embodiments of the present disclosure relate to multi-band FDD transceivers. An example transceiver includes a LO, configured to generate a LO signal to be shared between a receiver and a transmitter of the transceiver. Both the receiver and the transmitter use quadrature signal processing and are configured to multi-band operation. Sharing a single LO to perform frequency conversion of different frequency bands of received and transmitted signals advantageously allows reducing the number of LOs used in a multi-band FDD transceiver. 1. A multi-band frequency division duplex (FDD) transceiver system , comprising:a local oscillator (LO), configured to provide a LO signal;a receive (RX) quadrature mixer, configured to mix the LO signal with a received radio frequency (RF) signal to generate a downconverted RX signal, where the received RF signal includes a first received signal component in a first band of receiver frequencies and a second received signal component in a second band of receiver frequencies, the second band of receiver frequencies being separate from the first band of receiver frequencies; anda transmit (TX) quadrature mixer, configured to mix the LO signal with a TX signal to generate an upconverted RF TX signal, where the upconverted TX signal includes a first upconverted TX signal component in a first band of transmitter frequencies and a second upconverted TX signal component in a second band of transmitter frequencies, the second band of transmitter frequencies being separate from each one of the first band of transmitter frequencies, the first band of receiver frequencies, and the second band of receiver frequencies.2. The multi-band FDD transceiver system according to claim 1 , wherein a frequency of the LO signal is within the first band of receiver frequencies.3. The multi-band FDD transceiver system according to claim 1 , wherein a frequency of the LO signal is within the first band of transmitter frequencies.4. The multi-band FDD transceiver ...

METHOD AND ELECTRONIC DEVICE FOR PROCESSING INFORMATION

The present disclosure provides a method and electronic device for processing information. The method is applied in a solid state storage apparatus which is connected to an electronic device. The solid state storage apparatus supports N logical-address-to-physical-address mapping tables different from each other simultaneously, wherein N is an integer greater than or equal to 1. The method comprises: receiving identity information for a user from the electronic device; determining a first logical-address-to-physical-address mapping table corresponding to the user based on the identity information; and assigning the first logical-address-to-physical-address mapping table to the user. 1. A method for processing information , wherein the method is applied in a solid state storage apparatus which is connected to an electronic device , the solid state storage apparatus supports N logical-address-to-physical-address mapping tables different from each other simultaneously , wherein N is an integer greater than or equal to 1 , the method comprising:receiving identity information for a user from the electronic device;determining a first logical-address-to-physical-address mapping table corresponding to the user based on the identity information; andassigning the first logical-address-to-physical-address mapping table to the user.2. The method according to claim 1 , wherein the receiving identity information for a user from the electronic device comprises:receiving identity information for a user from the electronic device when the solid state storage apparatus is initialized.3. The method according to claim 1 , wherein the determining a first logical-address-to-physical-address mapping table corresponding to the user based on the identity information comprises: if yes, determining the first logical-address-to-physical-address mapping table; and', 'otherwise, creating a first logical-address-to-physical-address mapping table corresponding to the user., 'detecting whether the ...

CALCULATING DEVICE AND METHOD FOR A SPARSELY CONNECTED ARTIFICIAL NEURAL NETWORK

Aspects for modifying data in a multi-layer neural network (MNN) acceleration processor for neural networks are described herein. As an example, the aspects may include receiving one or more groups of input data and connection data. Further, the aspects may include modifying the one or more groups of input data based on the connection data. Further still, the aspects may include calculating one or more groups of output data based on the modified input data. 1. An apparatus for modifying data for neural networks , comprising: [ wherein the one or more groups of input data are stored as input elements in an input array and each of the input elements is identified by an input array index, and', 'wherein the one or more groups of output data are to be stored as output elements in an output array and each of the output elements is identified by an output array index;, 'receive one or more groups of input data for generating one or more groups of output data,'}, whether one of one or more weight values includes in a predetermined weight value array is designated for calculating a group of the output data to be stored as the output element identified by the corresponding output array index based on a group of the input data stored as the input element identified by the corresponding input array index, and', 'whether the weight value meets a predetermined condition; and, 'receive connection data that include one or more connection values, wherein each of the connection values corresponds to one of the input array indexes and one of the output array indexes and indicates'}, 'modify the received groups of input data based on the connection data; and, 'a data modifier configured to receive the modified input data from the data modifier and the predetermined weight value array; and', 'calculate the one or more groups of output data based on the modified input data and the weight values., 'a computing unit configured to2. The apparatus of claim 1 , wherein the predetermined ...

CALCULATING DEVICE AND METHOD FOR A SPARSELY CONNECTED ARTIFICIAL NEURAL NETWORK

Aspects for modifying data in a multi-layer neural network (MNN) acceleration processor for neural networks are described herein. As an example, the aspects may include receiving a predetermined weight value array and connection data. Further, the aspects may include modifying the weight values included in the predetermined weight value array based on the connection data. Further still, the aspects may include calculating one or more groups of output data based on the modified weight values. 1. An apparatus for modifying data for neural networks , comprising: receive one or more groups of input data, wherein the one or more groups of input data are stored as input elements in an input array and each of the input elements is identified by an input array index;', 'receive a predetermined weight value array that includes one or more weight values for calculating one or more groups of output data based on the one or more groups of input data, wherein the one or more groups of output data are to be stored as output elements in an output array and each of the output elements is identified by an output array index;', whether one of the weight values in the predetermined weight value array is designated for calculating a group of the output data to be stored as the output element identified by the corresponding output array index based on a group of the input data stored as the input element identified by the corresponding input array index, and', 'whether the weight value meets a predetermined condition; and, 'receive connection data that include one or more connection values, wherein each of the connection values corresponds to one of the input array indexes and one of the output array indexes and indicates'}, 'modify the weight values and the input data based on the connection data; and, 'a data modifier configured to receive the modified weight values and the modified input data from the data modifier; and', 'calculate the one or more groups of output data based on the ...

Calculating device and method for a sparsely connected artificial neural network

Aspects for modifying data in a multi-layer neural network (MNN) acceleration processor for neural networks are described herein. As an example, the aspects may include receiving one or more groups of input data, a predetermined weight value array, and connection data. Further, the aspects may include modifying the weight values included in the predetermined weight value array and the one or more groups of input data based on the connection data. Further still, the aspects may include calculating one or more groups of output data based on the modified weight values and the modified input data.

PROCESSING DATA BY USING SIMULTANEOUS MULTITHREADING

A computer implemented method and system for data processing. The method including: (a) setting at least one SMT preliminary value for at least one operating node; (b) monitoring performance metrics for the at least one operating node set to the at least one SMT preliminary value; and (c) determining a SMT revised value based on performance metrics. The system including: a memory; a processor communicatively coupled to the memory; and a feature selection module communicatively coupled to the memory and processor, wherein the feature selection module is configured to perform steps of a method including: setting, using a setting device, at least one SMT preliminary value for at least one operating node; monitoring, using a monitoring device, performance metrics for the at least one operating node set to the at least one SMT preliminary value; and determining, using a determining device, a SMT revised value based on performance metrics. 1. A computer implemented data processing method , comprising:(a) setting at least one simultaneous multithreading (SMT) preliminary value for at least one operating node;(b) monitoring performance metrics for the at least one operating node set to the at least one SMT preliminary value; and(c) determining a SMT revised value based on the performance metrics.2. The method according to claim 1 , wherein the method further comprises:(d) enabling the at least one operating node to operate under the SMT revised value.3. The method according to claim 1 , wherein:the at least one operating node operates under a MapReduce framework; andthe steps (a)-(c) are executed when the at least one operating node is in a first stage of the MapReduce framework.4. The method according to claim 3 , further comprising: only Map operation in the first stage and concurrent Map and Reduce operations in the second stage;', 'concurrent Map and Reduce operations in the first stage and only Reduce operation in the second stage; or', 'only Reduce operation in the ...

STANDARDIZED ENTITY REPRESENTATION LEARNING FOR SMART SUGGESTIONS

In an example, a plurality of user profiles in a social networking service are accessed. A heterogeneous graph structure comprising a plurality of nodes connected by edges is generated, each node corresponding to a different entity in the social networking service, each edge representing a co-occurrence of entities represented by nodes on each side of the edge in at least one of the user profiles. Weights are calculated for each edge of the heterogeneous graph structure, the weights being based on co-occurrence counts reflecting a number of user profiles in the plurality of user profiles in which corresponding nodes co-occurred. The heterogeneous graph structure is embedded into a d-dimensional space. A machine-learned model is then used to calculate a similarity score between a first node and second node by computing distance between the first node and the second node in the d-dimensional space. 1. A system comprising:one or more computer processors: and retrieve an indication of a plurality of different entities in a social networking service, including at least two entities having a different entity type,', 'access a plurality of user profiles in the social networking service;', 'generate a heterogeneous graph structure comprising a plurality of nodes connected by edges, each node corresponding to a different one of the entities in the plurality of different entities, each edge representing a co-occurrence of entities represented by nodes on each side of the edge m at least one of the user profiles;', 'compute weights for each edge of the heterogeneous graph structure, the weights being based on co-occurrence counts reflecting a number of user profiles in the plurality of user profiles in which corresponding nodes co-occurred;', 'embed the heterogeneous graph structure into a d-dimensional space;', 'use a machine-learned model to calculate a similarity score between a first node and a second node from the plurality of nodes by computing a distance between the ...

PERSONALIZED DEEP MODELS FOR SMART SUGGESTIONS RANKING

In an example, a deep learning network is used to calculate a similarity score between a first query in a social networking service and each of one or more suggestable entities in the social networking service. The suggestable entities are determined via a first machine learned model. The deep learning network takes as input the suggestable entities as well as a history of interactions with a graphical user interface of a social networking service by a first member of the social networking service, a history of queries performed via the graphical user interface by the first member, and the first query itself. 1. A system comprising:one or more processors; obtain a history of interactions with a graphical user interface of a social networking service by a first member of the social networking service;', 'obtain a history of queries performed via the graphical user interface by the first member;', 'receive a first query initiated by the first member, the first query identifying a first set of one or more entities in the social networking service;', 'form a query facet entities vector from the first set of the one or more entities;', 'submit the query facet entities vector to a first machine learned model to identity a set of one or more suggestable entities, each of the one or more suggestable entities being an entity similar to, but not identical to, an entity in the first set of the one or more entities, based on embeddings learned by the first machine learned model;', 'form a set of suggestable entities vectors from the set of the one or more suggestable entities;', 'form a set of history of interactions vectors from the history of interactions;', 'form a set of history of queries vectors from the history of queries;', 'pass the query facet entities vector, the set of suggestable entities vectors, the set of history of interactions vectors, and the set of history of queries vectors into a deep learning network to calculate a similarity score between the first query ...

JOINT REPRESENTATION LEARNING OF STANDARDIZED ENTITIES AND QUERIES

An indication of a plurality of different entities in a social networking service is received, including at least two entities having a different entity type. A plurality of user profiles in the social networking service is accessed. A first machine-learned model is used to learn embeddings for the plurality of different entities in a d-dimensional space. A second machine-learned model is used to learn an embedding for each of one or more query terms that are not contained in the indication of the plurality of different entities in the social networking service, using the embeddings for the plurality of different entities learned using the first machine-learned model, the second-machine learned model being a deep structured semantic model (DSSM). A similarity score between a query term and an entity is calculated by computing distance between the embedding for the query term and the embedding for the entity in the d-dimensional space. 1. A system comprising:one or more processors; and retrieve an indication of a plurality of different entities in an online system, including at least two entities having a different entity type;', 'use a first machine-learned model to learn embeddings for the plurality of different entities in a d-dimensional space;', 'use a second machine-learned model to learn an embedding for each of one or more query terms that are not contained in the indication of the plurality of different entities in the social networking service, using the embeddings for the plurality of different entities learned using the first machine-learned model, the second-machine learned model being a deep structured semantic model (DSSM); and', 'calculate a similarity score between a query term and an entity by computing a distance between the embedding for the query term and the embedding for the entity in the d-dimensional space., 'a computer-readable medium having instructions stored thereon, which, when executed by the one or more processors, cause the system to2 ...

LIGHT WEIGHT INDEX FOR QUERYING LOW-FREQUENCY DATA IN A BIG DATA ENVIRONMENT

The present disclosure relates to searching for and committing low-frequency data to a database. An example method generally includes receiving, from a requesting application, a query for data from the data repository. A database system retrieves a set of indices associated with the data specified in the query from an index table in the data repository. Upon determining that the set of indices comprises a non-null set, the database system retrieves records associated with each index in the set of indices from a data table associated with the index table and returns the retrieved records to the requesting application. 1. A method for efficiently querying for low-frequency data in a data repository , comprising:receiving, from a requesting application, a query for data from a data repository;retrieving a set of indices associated with the data specified in the query from an index data table in the data repository;upon determining that the set of indices comprises a non-null set, retrieving records associated with each index in the set of indices from a data table in the data repository associated with the index data table; andreturning the retrieved records to the requesting application.2. The method of claim 1 , further comprising: upon determining that the set of indices comprises a null set claim 1 , executing the query against the data table.3. The method of claim 1 , further comprising:receiving a request to write data to the data table, wherein the request includes information associated with a monitored data entry in the data table;writing a record to the data table including the information;determining if the information is associated with a second non-null set of indices in the index data table; andupon determining that the information is associated with the second non-null set of indices in the index data table, adding an index of the written record to the set of indices in the index data table.4. The method of claim 3 , further comprising:calculating a ...

METHOD AND DEVICE FOR PROCESSING RANDOM ACCESS PREAMBLE

Provided are a method and device for processing a random access preamble. The method includes: determining, according to a coverage radius of a cell, a guard time (GT) that is needed when a user equipment (UE) communicates with a base station, wherein the coverage radius of the cell is greater than 100 kilometres; and, determining, according to the GT, a random access preamble transmitted by the UE to the base station, wherein the random access preamble is transmitted on a special subframe and at least one regular subframe of a radio frame when the UE communicates with the base station. The solution solves the problem that the maximum coverage range of a single-station cell cannot satisfy a need in a special scenario, further increases the coverage radius of the single-station cell, and satisfies the need in the special scenario for an extra-large radius in cell coverage range. 1. A method for processing a random access preamble , comprising:determining, according to a coverage radius of a cell, a guard time (GT) that is needed when a user equipment (UE) communicates with a base station, wherein the coverage radius of the cell is greater than 100 kilometres; anddetermining, according to the GT, a random access preamble transmitted by the UE to the base station, wherein the random access preamble is transmitted on a special subframe and at least one regular subframe of a radio frame when the UE communicates with the base station.2. The method according to claim 1 , wherein before determining claim 1 , according to the GT claim 1 , the random access preamble transmitted by the UE to the base station claim 1 , the method further comprises:determining, according to the coverage radius of the cell, a guard period (GP) that is needed when the UE communicates with the base station.3. The method according to claim 1 , wherein determining claim 1 , according to the GT claim 1 , the random access preamble transmitted by the UE to the base station comprises:extending the ...

DRESSING METHOD, DEVICE, COMPUTER READABLE MEDIUM, AND ELECTRONIC DEVICE

A costume replacement method includes, based on a user input, obtaining information of a model including a plurality of parts, the plurality of parts including a target part of which a costume is to be replaced, obtaining image set object data corresponding to the model of which the information is obtained, the image set object data including a plurality of pieces of image data respectively corresponding to the plurality of parts, and obtaining costume replacement data corresponding to the target part included in the model of which the information is obtained. The method further includes updating the obtained image set object data by updating one of the plurality of pieces of image data corresponding to the target part, based on the obtained costume replacement data, and rendering the model, based on the updated image set object data, to update the costume of the target part included in the model. 1. A costume replacement method , the costume replacement method being performed by at least one processor , and the costume replacement method comprising:based on a user input, obtaining information of a model comprising a plurality of parts, the plurality of parts comprising a target part of which a costume is to be replaced;obtaining image set object data corresponding to the model of which the information is obtained, the image set object data comprising a plurality of pieces of image data respectively corresponding to the plurality of parts;obtaining costume replacement data corresponding to the target part comprised in the model of which the information is obtained;updating the obtained image set object data by updating one of the plurality of pieces of image data corresponding to the target part, based on the obtained costume replacement data; andrendering the model, based on the updated image set object data, to update the costume of the target part comprised in the model.2. The costume replacement method according to claim 1 , wherein the updating the obtained ...

APPARATUS AND METHOD FOR PERFORMING A FORWARD OPERATION OF ARTIFICIAL NEURAL NETWORKS

Aspects for executing forward propagation of artificial neural network are described here. As an example, the aspects may include a plurality of computation modules connected via an interconnection unit; and a controller unit configured to decode an instruction into one or more groups of micro-instructions, wherein the plurality of computation modules are configured to perform respective groups of the micro-instructions. 1. An apparatus for forward propagation of a multilayer neural network , comprising:one or more slave computation modules configured to parallelly calculate groups of slave output values based on an input vector received via an interconnection unit; and calculate a merged intermediate vector based on one or more intermediate result vectors combined by the interconnection unit based on the multiple groups of slave output values calculated by the multiple slave computation modules, and', 'generate an output vector based on the merged intermediate vector., 'a master computation module configured to2. The apparatus of claim 1 , further comprising a controller unit configured to decode an instruction into one or more groups of micro-instructions that instruct the master computation module and the one or more slave computation modules to perform one or more operations.3. The apparatus of claim 1 , wherein the master computation module is configured to perform one operation selected from the group consisting of:adding a bias value to the merged intermediate vector;activating the merged intermediate vector with an activation function;outputting a predetermined value based on a comparison between the merged intermediate vector and a random number; andpooling the merged intermediate vector.4. The apparatus of claim 1 , wherein each of the slave computation modules includes a slave neuron caching unit configured to store the input vector.5. The apparatus of claim 1 , wherein the interconnection unit is structured as a binary tree including one or more levels ...

APPARATUS AND METHOD FOR EXECUTING REVERSAL TRAINING OF ARTIFICIAL NEURAL NETWORK

An apparatus for executing backpropagation of an artificial neural network comprises an instruction caching unit, a controller unit, a direct memory access unit, an interconnection unit, a master computation module, and multiple slave computation modules. For each layer in a multilayer neural network, weighted summation may be performed on input gradient vectors to calculate an output gradient vector of this layer. The output gradient vector may be multiplied by a derivative value of a next-layer activation function on which forward operation is performed, so that a next-layer input gradient vector can be obtained. The input gradient vector may be multiplied by an input neuron counterpoint in forward operation to obtain the gradient of a weight value of this layer, and the weight value of this layer can be updated according to the gradient of the obtained weight value of this layer. 1. An apparatus for backpropagation of a multilayer neural network , comprising: receive a first output gradient vector from an adjacent layer, and', 'respectively multiply each element of the first output gradient vector with a derivative of an activation function of a current layer to obtain an input gradient vector; and, 'a master computation module configured to parallelly calculate gradients of weight values based on the input gradient vector, and', 'update the weight values based on the respectively calculated gradients., 'multiple slave computation modules connected to the master computation module via an interconnection unit, wherein the multiple slave computation modules are configured to2. The apparatus of claim 1 , wherein the multiple slave computation modules are configured to parallelly calculate portions of a second output gradient vector based on the input gradient vector calculated by the master computation module.3. The apparatus of claim 2 , wherein the interconnection unit is configured to combine the portions of the second output gradient vector to generate the ...

TASK SCHEDULING METHOD AND APPARATUS ON HETEROGENEOUS MULTI-CORE RECONFIGURABLE COMPUTING PLATFORM

Embodiments of the present invention provide a task scheduling method and apparatus on a heterogeneous multi-core reconfigurable computing platform. The method includes: when determining that a to-be-executed hardware task is in a ready state, adding the to-be-executed hardware task into a target hardware task queue corresponding to a function of the to-be-executed hardware task; reconfiguring, according to a priority of the to-be-executed hardware task and a usage status of multiple reconfigurable resource packages, at least one reconfigurable resource package in the multiple reconfigurable resource packages into a target intellectual property IP core that can execute the to-be-executed hardware task. According to the embodiments of the present invention, a more proper resource reconfiguration occasion can be determined, and a more appropriate reconfigurable resource package can be used for resource reconfiguration, thereby reducing reconfiguration overheads and improving utilization of reconfigurable resources in a task scheduling process. 1. A task scheduling method on a heterogeneous multi-core reconfigurable computing platform , wherein the heterogeneous multi-core reconfigurable computing platform comprises multiple reconfigurable resource packages , and the method comprises:when determining that a to-be-executed hardware task is in a ready state, adding the to-be-executed hardware task into a target hardware task queue corresponding to a function of the to-be-executed hardware task;reconfiguring, according to a priority of the to-be-executed hardware task and a usage status of the multiple reconfigurable resource packages, at least one reconfigurable resource package in the multiple reconfigurable resource packages into a target intellectual property(IP)core that can execute the to-be-executed hardware task, wherein the priority denotes an execution order of the hardware task; andexecuting the hardware task in the target hardware task queue by using the ...

COMPUTERIZED RENDERING OF ANIMATED OBJECTS USING POLYNOMIAL PARTICLE-IN-CELL METHOD

A method of rendering an animated object includes: (1) determining momentums of a plurality of particles of the object as sums of polynomials; (2) transferring the momentums of the particles of the object to a grid including a plurality of grid nodes; (3) updating momentums of the grid nodes based on the transferred momentums of the particles; (4) transferring the updated momentums of the grid nodes to the particles of the object; (5) updating positions of the particles based on the updated momentums of the grid nodes; and (6) outputting a visualization of the object based on the updated positions of the particles of the object. 1. A method of rendering an animated object , comprising:determining momentums of a plurality of particles of the object as sums of polynomials;transferring the momentums of the particles of the object to a grid including a plurality of grid nodes;updating momentums of the grid nodes based on the transferred momentums of the particles;transferring the updated momentums of the grid nodes to the particles of the object;updating positions of the particles based on the updated momentums of the grid nodes; andoutputting a visualization of the object based on the updated positions of the particles of the object.2. The method of claim 1 , wherein the polynomials include a plurality of component-wise velocity modes.3. The method of claim 2 , wherein the component-wise velocity modes include a constant velocity mode.4. The method of claim 2 , wherein the component-wise velocity modes include a linear velocity mode.5. The method of claim 2 , wherein the component-wise velocity modes include a multi-linear velocity mode.6. The method of claim 2 , wherein the component-wise velocity modes include a multi-quadratic velocity mode.7. The method of claim 1 , wherein kinetic energies are conserved when the momentums of the particles of the object are transferred to the grid.8. The method of claim 1 , wherein the polynomials having a polynomial basis which is ...

LITHIUM ION BATTERY USING NON-AQUEOUS ELECTROLYTE

Disclosed is a lithium ion battery using a non-aqueous electrolyte. The lithium ion battery comprises a positive electrode, a negative electrode, a separator arranged between the positive electrode and the negative electrode, and a non-aqueous electrolyte of the lithium ion battery. An active substance of the positive electrode comprises LiFePO; the non-aqueous electrolyte of the lithium ion battery comprises a non-aqueous organic solvent, a lithium salt and an additive; and the additive at least comprises (A) vinylene carbonate, and at the same time, also comprises at least one of (B) a compound shown in a structural formula 1 and (C) a fluorobenzene, wherein n is a natural number of 1-3, and each of R, R, Rand Ris independently selected from one of a hydrogen atom, a fluorine atom, and an alkyl group with 1-6 carbon atoms. The lithium ion battery has a long cycle life, and at the same time, the battery has excellent high and low temperature performance. 2. The lithium-ion battery according to claim 1 , wherein the additive (A) accounts for 0.2% to 5% claim 1 , preferably 0.5% to 3% of the total weight of the electrolyte.3. The lithium-ion battery according to claim 1 , wherein the additive (B) accounts for 0.1% to 5% claim 1 , preferably 0.5% to 3% of the total weight of the electrolyte.4. The lithium-ion battery according to claim 1 , wherein the additive (C) accounts for 0 claim 1 , 1% to 20% claim 1 , preferably 1% to 10% of the total weight of the electrolyte.5. The lithium-ion battery according to claim 1 , wherein the compound represented by Structural Formula 1 is ethylene sulfate or 1 claim 1 ,3-propanediol sulfate.6. The lithium-ion battery according to claim 1 , wherein the non-aqueous organic solvent is selected from one or more of ethyl methyl carbonate claim 1 , dimethyl carbonate claim 1 , diethyl carbonate claim 1 , ethylene carbonate claim 1 , propylene carbonate claim 1 , butylene carbonate claim 1 , ethylene sulfite claim 1 , propylene sulfite ...

NEURAL NETWORK CONVOLUTION COMPUTATION METHOD AND DEVICE, AND COMPUTER-READABLE STORAGE MEDIUM

Aspects of a neural network convolution device are described herein. The aspects may include a matrix transformer and a matrix multiplication module. The matrix transformer may be configured to receive an input data matrix and a weight matrix, transform the input data matrix into a transformed input data matrix based on a first transformation matrix, and transform the weight matrix into a transformed weight matrix based on a second transformation matrix. The matrix multiplication module may be configured to multiply one or more input data elements in the transformed input data matrix with one or more weight elements in the transformed weight matrix to generate an intermediate output matrix. The matrix transformer may be further configured to transform the intermediate output matrix into an output matrix based on an inverse transformation matrix. 1. A neural network convolution device , comprising: receive an input data matrix and a weight matrix,', 'transform the input data matrix into a transformed input data matrix based on a first transformation matrix, and', 'transform the weight matrix into a transformed weight matrix based on a second transformation matrix; and, 'a matrix transformer configured to'}a matrix multiplication module configured to multiply one or more input data elements in the transformed input data matrix with one or more weight elements in the transformed weight matrix to generate an intermediate output matrix, wherein the matrix transformer is further configured to transform the intermediate output matrix into an output matrix based on an inverse transformation matrix.2. The neural network convolution device of claim 1 , wherein the first transformation matrix claim 1 , the second transformation matrix claim 1 , and the inverse transformation matrix are predetermined based on a first size of the input data matrix claim 1 , a second size of the weight matrix claim 1 , and a stride for moving a convolution kernel.3. The neural network convolution ...

MULTIPLICATION AND ADDITION DEVICE FOR MATRICES, NEURAL NETWORK COMPUTING DEVICE, AND METHOD

Aspects of a neural network operation device are described herein. The aspects may include a matrix element storage module configured to receive a first matrix that includes one or more first values, each of the first values being represented in a sequence that includes one or more bits. The matrix element storage module may be further configured to respectively store the one or more bits in one or more storage spaces in accordance with positions of the bits in the sequence. The aspects may further include a numeric operation module configured to calculate an intermediate result for each storage space based on one or more second values in a second matrix and an accumulation module configured to sum the intermediate results to generate an output value. 1. A neural network operation device , comprising: receive a first matrix that includes one or more first values, wherein each of the one or more first values is represented as a sequence that includes one or more bits, and', 'respectively store the one or more bits in one or more storage spaces in accordance with positions of the one or more bits in the sequence;, 'a matrix element storage module configured toa numeric operation module configured to calculate an intermediate result for each storage space based on one or more second values in a second matrix, wherein the one or more second values correspond to non-zero values stored in the storage space; andan accumulation module configured to sum the intermediate results to generate an output value.2. The neural network operation device of claim 1 , further comprising a submatrix dividing module configured to select a portion of an input data matrix as the second matrix.3. The neural network operation device of claim 1 , wherein the one or more bits that represent each of the first values include a sign bit.4. The neural network operation device of claim 3 , wherein the sign bits of the first values are stored in a sign storage space.5. The neural network operation ...

APPARATUS AND METHOD FOR PERFORMING A FORWARD OPERATION OF ARTIFICIL NEURAL NETWORKS

Aspects for executing forward propagation of artificial neural network are described here. As an example, the aspects may include a plurality of computation modules connected via an interconnection unit; and a controller unit configured to decode an instruction into one or more groups of micro-instructions, wherein the plurality of computation modules are configured to perform respective groups of the micro-instructions. 1. An apparatus for forward propagation of a multilayer neural network , comprising:a plurality of computation modules connected via an interconnection unit; anda controller unit configured to decode an instruction into one or more groups of micro-instructions, wherein the plurality of computation modules are configured to perform respective groups of the micro-instructions.2. The apparatus of claim 1 , wherein the plurality of computation modules further comprise:multiple slave computation modules configured to parallelly calculate multiple groups of slave output values based on an input vector received via the interconnection unit; and to calculate a merged intermediate vector based on one or more intermediate result vectors combined by the interconnection unit based on the multiple groups of slave output values calculated by the multiple slave computation modules, and', 'to generate an output vector based on the merged intermediate vector., 'a master computation module configured'}3. The apparatus of claim 2 , wherein the master computation module is configured to perform one operation selected from the group consisting of:adding a bias value to the merged intermediate vector;activating the merged intermediate vector with an activation function;outputting a predetermined value based on a comparison between the merged intermediate vector and a random number; andpooling the merged intermediate vector.4. The apparatus of claim 2 , wherein each of the slave computation modules includes a slave neuron caching unit configured to store the input vector. ...

APPARATUS AND METHOD FOR EXECUTING REVERSAL TRAINING OF ARTIFICIAL NEURAL NETWORK

An apparatus for executing backpropagation of an artificial neural network comprises an instruction caching unit, a controller unit, a direct memory access unit, an interconnection unit, a master computation module, and multiple slave computation modules. For each layer in a multilayer neural network, weighted summation may be performed on input gradient vectors to calculate an output gradient vector of this layer. The output gradient vector may be multiplied by a derivative value of a next-layer activation function on which forward operation is performed, so that a next-layer input gradient vector can be obtained. The input gradient vector may be multiplied by an input neuron counterpoint in forward operation to obtain the gradient of a weight value of this layer, and the weight value of this layer can be updated according to the gradient of the obtained weight value of this layer. 1. An apparatus for backpropagation of a multilayer neural network , comprising:a plurality of computation modules connected via an interconnection unit; anda controller unit configured to decode an instruction into one or more groups of micro-instructions, wherein the one or more groups of micro-instructions respectively control operations of one or more groups of the plurality of computation modules.2. The apparatus of claim 1 , wherein the plurality of computation modules further comprise:a master computation module configured to calculate an input gradient vector based on a first output gradient vector from an adjacent layer; andmultiple slave computation modules configured to parallelly calculate portions of a second output gradient vector based on the input gradient vectors calculated by the master computation module.3. The apparatus of claim 2 , wherein the interconnection unit is configured to combine the portions of the second output gradient vector to generate the second output gradient vector.4. The apparatus of claim 1 , further comprising:an instruction caching unit ...

APPARATUS AND METHOD FOR COMPRESSION CODING FOR ARTIFICIAL NEURAL NETWORK

A compression coding apparatus for artificial neural network, including memory interface unit, instruction cache, controller unit and computing unit, wherein the computing unit is configured to perform corresponding operation to data from the memory interface unit according to instructions of controller unit; the computing unit mainly performs three steps operation: step one is to multiply input neuron by weight data; step two is to perform adder tree computing and add the weighted output neuron obtained in step one level-by-level via adder tree, or add bias to output neuron to get biased output neuron; step three is to perform activation function operation to get final output neuron. The present disclosure also provides a method for compression coding of multi-layer neural network. 1. A neural network processor , comprising: receive one or more first weight values of a first bit length and first input neuron data, and', 'convert the one or more first weight values to one or more second weight values of a second bit length, wherein the second bit length is less than the first bit length; and, 'a floating-point number converter configured to'} receive the first input neuron data from the neuron data cache,', 'calculate first output neuron data based on the first input neuron data and the second weight values, and', 'calculate one or more weight gradients to update the one or more first weight values., 'a computing unit configured to'}2. The neural network processor of claim 1 , further comprising:a weight cache configured to store the one or more first weight values, anda neuron data cache configured to store the first input neuron data.3. The neural network processor of claim 1 , wherein the computing unit includes:one or more multipliers configured to multiply a first portion of the first neuron data by a second portion of the first neuron data;one or more adders configured to add multiplication results output from the one or more multipliers; andone or more ...