Design method of onboard computer board-level low-direct-current impedance coplanar electromagnetic band-gap power supply layer
Technical Field The invention relates to the technical field of power supply board design, in particular to an airborne computer board-level low-direct-current impedance coplanar electromagnetic bandgap power layer design method. Background Art With the increase in digital circuit rate, clock frequency and PCB design complexity, related power integrity issues are increasingly prominent. Compared with a common electronic device, the airborne computer has a more severe working environment and a higher performance standard, so that higher requirements are also brought to the integrity of the power supply. PCB power integrity issues mainly relate to voltage drop and signal noise. Since these two parts are negatively affected to a certain extent, the design of the power distribution network is critical to achieve an overall optimum balance. Existing power distribution network designs still adopt traditional methods, namely designers rely on experience and intuition to perform initial design, and the initial design is corrected through multiple experiments. The design efficiency is low, the performance of the power supply board is prevented from being improved, and the design requirement cannot be met. Content of the invention In order to overcome the defects of the prior art, the invention aims to provide an airborne computer board-level low-direct-current impedance coplanar electromagnetic bandgap power layer design method which greatly improves the design efficiency while improving the comprehensive design performance of the power supply layer. In order to achieve the above object, the technical scheme adopted by the invention is as follows. The invention discloses an airborne computer board-level low-direct-current impedance coplanar electromagnetic bandgap power layer design method. 1)Construction of an equivalent model: The selected electromagnetic bandgap unit is equivalent to have a conductivity σ.0 A metal patch of the metal patch is used as a material of a low-conductivity substrate in a model, and has high conductivity σ.p The copper power distribution channel is formed by the growth of the upper part of the base material; meanwhile, an effective assumption of uniform and uniform seepage of current on the whole surface of the equivalent metal paster is made. 2)A finite element model of a design domain, namely a base structure, is established according to the material parameters of an equivalent metal patch, and voltage and current load boundary conditions are applied to the basic structure according to actual voltage and current input conditions of the power supply board. The power supply board design domain is rectangular in 45mm×60mm, and the design domain is insulated, wherein one long side midpoint is a current input point, the voltage is 1 V, the input current is 1A, and the design domain is uniform in osmotic electricity. 1mm×1mm 3)The structure of the flexible growth unit is that the final configuration of the power distribution network is composed of a plurality of flexible growth units, and the flexible growth unit is expressed explicitly by taking the zero-level set of the horizontal set function. The horizontal set function is a horizontal set function. Where. (X)i Y-yi Is the coordinate of the point A of the flexible growth unit, L is the unit half length, θ is the unit inclination angle, t1 T-t2 And t3 The semi-widths of the flexible growth units A, B, C are respectively, 7 of the variables representing the geometrical parameters of the flexible growth unit can define a flexible growth unit. Xi ＝i Y-yi , L, t1 T-t2 T-t3 ] Θ]T The node of any one coordinate on the base structure is (x, y), and the i th flexible growth unit may output a value φ of a corresponding horizontal set function.i The final horizontal set function value of the node takes the maximum value φ of the resulting values.s (X, y) ＝ max (φ)1 PHφ2 PHφ3 , φn , N is the number of flexible growth units. 4)Finite Element Analysis: After obtaining the horizontal set function value, the conductivity of each quadrilateral shell unit on the base structure can be interpolated by the conductivity of the four nodes, and the finite element method has the finite element method. Here K is a conductivity matrix of the entire structure. It is the voltage, J is the current load, so far the finite element model of the flexible growing unit is obtained. 5)Adaptive growth method: 5.1) Establishment of a problem mathematical model. Objective function: voltage drop Minimum, average voltage is maximum. Design variable: Xi ＝1 , X1 . Xn ]. Constraints: V ≤ VMax X-ray and X-rayi Is the geometric parameter of i th flexible growth unit, V is the total amount of material used for the high conductivity material, V.Max This is the maximum amount of material allowed. 5.2) The growth process is divided into a main pulse and a secondary pulse, the growth of the subpulse is started after the main pulse growth is completed, and the growth competition and the local reconstruction of the two growth strategies are introduced. 5.2.1) Power distribution network parameter initialization: After establishing the design domain, the maximum material usage β of the power distribution network is set.0 Β-cyclodextrin for each step in the course of growth(k) The conductivity of the base structure and the power distribution network is set to σ, respectively.0 And σp Setting initial value X of design variable0 Minimum XMin Maximum XMax . 5.2.2) Growth competition: growth competition is the process of growing competition, i.e. the growth of the flexible growth unit to obtain optimal geometric parameters; the local optimization is a process of optimizing the width of all the flexible growth units; and the global optimization is the process of optimizing the width of all the flexible growth units. 5.2.3) Local reconstruction: configuration reconstruction takes place in the growing process, is used for deciding whether the terminal of certain flexible growth unit continues to grow out the next stage unit. Setting two thresholds: growth threshold Wb And degradation threshold Wd : Newly grown unit intermediate width value t2 ＞ Wb , Then the newly grown cell will be retained; when t2 ＜ Wd , Newly grown cells will be removed; when W is usedd ≤ t2 ≤ Wb The newly grown unit will be retained, but the end thereof cannot grow to the next stage unit anymore. Growth race and configuration reconstruction continues iterate with MMA algorithm updates until the amount of material reaches the maximum β set at initialization.0 , An optimized design of an airborne computer board-level low-direct-current impedance coplanar electromagnetic bandgap power distribution network formed by the growth of the flexible growth unit is obtained. The beneficial effect of the invention is as follows: To the flexible growth design method, performance parameters of the power supply layer can be obtained through a finite element method in the design stage without relying on design experience of designers, design reliability is improved, better design results can be obtained compared with traditional design, and design cost is reduced. Description of drawings 1 Is a flowchart of the method of the present invention. 2 Is a schematic diagram of a construction process of an equivalent model of the present invention. 3 Is an initialization schematic diagram of a finite element model of the basic structure of the present invention. 4 Is a schematic view of a flexible growth unit of the present invention. 5 Is a schematic diagram of a combination method of a flexible growth unit and a finite element analysis of the present invention. 6 Is a schematic diagram of the growth process of the adaptive growth algorithm of the present invention. Mode of execution The present invention will be further described in detail with reference to the accompanying drawings and embodiments. To FIG. 1, an airborne computer board-level low direct-current impedance coplanar electromagnetic bandgap power layer design method includes the following steps. 1)Construction of an equivalent model: The selected electromagnetic bandgap unit is equivalent to have a conductivity σ.0 A metal patch of the metal patch is used as a material of a low-conductivity substrate in a model, and has high conductivity σ.p The copper power distribution channel is formed by the growth of the upper part of the base material; meanwhile, an effective assumption of uniform and uniform seepage of current on the whole surface of the equivalent metal paster is made. As shown 2, for a certain EBG structure, a rectangular metal patch having the same external dimension but lower conductivity can be equivalently used, and the equivalent method is as follows: A resistance R between two points A and B in EBG structure and a conductivity σ thereof are obtained.p The relation formula between the equivalent metal patches A and B is determined, and the electrical conductivity σ and the electrical conductivity thereof are calculated between the two points of the equivalent metal patches A and B.0 The equation of the relation between the equivalent metal patches σ and the conductivity σ corresponding to the equivalent metal patches can be obtained.0 . 2)The finite element model of the design domain is established by the quadrilateral shell unit, and voltage and current load boundary conditions are applied to the basic structure according to actual voltage and current input conditions of the power supply plate. As shown 3, the power panel design domain is rectangular in 45mm×60mm, and the design domain is insulated, wherein one long side midpoint is a current input point where the voltage is 1 V, the input current is 1A, and the design domain is uniform in osmotic electricity. 1mm×1mm 3)The structure of the flexible growth unit: the final configuration of the power distribution network is composed of a plurality of flexible growth units, and the flexible growth unit is expressed explicitly by taking the zero-level set of the horizontal set function. As shown 4, the horizontal set function is a horizontal set function. Where. (X)i Y-yi Is the coordinate of the point A of the flexible growth unit, L is the unit half length, θ is the unit inclination angle, t1 T-t2 And t3 The semi-widths of the flexible growth units A, B, C three points, respectively, 7 represent the variables of the geometrical parameters of the flexible growth unit, which may define a flexible growth unit. Xi ＝i Y-yi , L, t1 T-t2 T-t3 ] Θ]T The node of any one coordinate on the base structure is (x, y), and the i th flexible growth unit may output a value φ of a corresponding horizontal set function.i The final horizontal set function value of the node takes the maximum value φ of the resulting values.s (X, y) ＝ max (φ)1 PHφ2 PHφ3 , φn , N is the number of flexible growth units. 4)Finite Element Analysis: After obtaining the level set function values, the conductivity of each shell unit on the base structure may be interpolated from the conductivity of the four nodes. H (x) is Heaviavir function. Is a horizontal set function at the i-th node of the base structure unit e. This embodiment employs Heaviavir functions in the following form. Thus, the finite element method is obtained. Here K is a conductivity matrix of the entire structure. Is the voltage and J is the current load. So far, the finite element model of the flexible growing unit can be obtained as shown 5. 5)Adaptive growth method: 5.1) Establishment of a problem mathematical model. Design variables are geometrical parameters of each flexible growth unit. The average voltage in the design domain is. Wherein N is the number of nodes of the lattice of the base structure. The constraint conditions for the maximum material usage are expressed as follows. The material usage constraint condition of each newly grown flexible growth unit is expressed as follows. In conclusion, the mathematical model of power distribution network topology optimization is: K Ω 1, 2, k)Max I aopipelineao1 , 2, 3, n(k) S.t. XMin ≤ Xi(k) ≤ XMax 5.2) The growth process is divided into a main pulse and a secondary pulse, the growth of the subpulse is started after the main pulse growth is completed, and the growth competition and local reconstruction of the two growth strategies are introduced as shown 6. 5.2.1) Power distribution network parameter initialization: After establishing the design domain, the maximum material usage β of the power distribution network is set.0 Β-cyclodextrin for each step in the course of growth(k) The conductivity of the base structure and the power distribution network is set to σ, respectively.0 And σp Setting initial value X of design variable0 Minimum XMin Maximum XMax . 5.2.2) Growth competition: growth competition is the process of growing competition, i.e. the growth of the flexible growth unit to obtain optimal geometrical parameters; the local optimization is optimization of a set of geometrical parameters for each flexible growth unit; the global optimization is a process of optimizing the width of all the flexible growth units. 5.2.3) Local reconstruction: configuration reconstruction takes place in the growing process, is used for deciding whether the terminal of certain flexible growth unit continues to grow out the next stage unit. Setting two thresholds: growth threshold Wb And degradation threshold Wd : Newly grown unit intermediate width value t2 ＞ Wb , Then the newly grown cell will be retained; when t2 ＜ Wd , Newly grown cells will be removed; when W is usedd ≤ t2 ≤ Wb The newly grown unit will be retained, but the end thereof cannot grow to the next stage unit anymore. Growth competition and configuration reconstruction is iterated with MMA algorithm updates until the amount of material reaches a maximum β of the initial setting.0 So far, the optimal design of the airborne computer board-level low-direct-current impedance coplanar electromagnetic bandgap power distribution network formed by the growth of the flexible growth unit is obtained. Simulation and experimental verification of the validity of the method of the present invention are described below. In order to explain the effective role of the method in the design of the power supply board, the distribution network formed by the flexible growth of the traditional configuration distribution network and the method disclosed by the invention is respectively combined with two different structures EBG to form a common four power supply board. 1, Voltage drop and current density comparison: As can be seen from the above result, both the voltage and the current, the configuration formed by the flexible growth has better performance and the comprehensive performance is higher than about 30% of the conventional configuration. 2, Noise suppression effect comparison: To first kinds EBG structure power supply boards, the traditional configuration and the growth formation configuration noise suppression capability are similar, noise suppression capability of the power supply board composed of 2GHz - 20GHz EBG is lower than - 2525dB, noise suppression capability of the formed configuration is obviously higher than that of a traditional configuration, and noise can be 1.7 GHz - 19.5 GHz suppressed in - 2525dB in second range. In conclusion, the flexible growth design method has higher design efficiency, and the designed power supply board has better performance. A finite element analysis is carried out, then the finite element analysis is carried out, and then self-adaptive growth, growth competition and configuration reconstruction are performed continuously under the update of MMA algorithm until the material consumption reaches the maximum value β set in initialization.0 Compared with a traditional design, the design method has higher design efficiency, better design results can be obtained, and design cost is reduced. 1.An airborne computer board-level low-direct-current impedance coplanar electromagnetic bandgap power layer design method is characterized by comprising the following steps: 1) Construction of an equivalent model: The selected electromagnetic bandgap unit is equivalent to have a conductivity σ.0 A metal patch of the metal patch is used as a material of a low-conductivity substrate in a model, and has high conductivity σ.p The copper power distribution channel is formed by the growth of the upper part of the base material; meanwhile, an effective assumption of uniform and uniform seepage of current on the whole surface of the equivalent metal paster is made. 2)A finite element model of a design domain, namely a base structure, is established according to the material parameters of an equivalent metal patch, and voltage and current load boundary conditions are applied to the basic structure according to actual voltage and current input conditions of the power supply board. The power supply board design domain is rectangular in 45mm×60mm, and the design domain is insulated, wherein one long side midpoint is a current input point, the voltage is 1 V, the input current is 1A, and the design domain is uniform in osmotic electricity. 1mm×1mm 3)The structure of the flexible growth unit is that the final configuration of the power distribution network is composed of a plurality of flexible growth units, and the flexible growth unit is expressed explicitly by taking the zero-level set of the horizontal set function. The horizontal set function is a horizontal set function. Where. (X)i Y-yi Is the coordinate of the point A of the flexible growth unit, L is the unit half length, θ is the unit inclination angle, t1 T-t2 And t3 The semi-widths of the flexible growth units A, B, C are respectively, 7 of the variables representing the geometrical parameters of the flexible growth unit can define a flexible growth unit. Xi ＝i Y-yi , L, t1 T-t2 T-t3 ] Θ]T The node of any one coordinate on the base structure is (x, y), and the i th flexible growth unit may output a value φ of a corresponding horizontal set function.i The final horizontal set function value of the node takes the maximum value φ of the resulting values.s (X, y) ＝ max (φ)1 PHφ2 PHφ3 , φn , N is the number of flexible growth units. 4)Finite Element Analysis: After obtaining the horizontal set function value, the conductivity of each quadrilateral shell unit on the base structure can be interpolated by the conductivity of the four nodes, and the finite element method has the finite element method. Here K is a conductivity matrix of the entire structure. It is the voltage, J is the current load, so far the finite element model of the flexible growing unit is obtained. 5)Adaptive growth method: 5.1) Establishment of a problem mathematical model. Objective function: voltage drop Minimum, average voltage is maximum. Design variable: Xi ＝1 , X1 . Xn ]. Constraints: V ≤ VMax X-ray and X-rayi Is the geometric parameter of i th flexible growth unit, V is the total amount of material used for the high conductivity material, V.Max This is the maximum amount of material allowed. 5.2) The growth process is divided into a main pulse and a secondary pulse, the growth of the subpulse is started after the main pulse growth is completed, and the growth competition and the local reconstruction of the two growth strategies are introduced. 5.2.1) Power distribution network parameter initialization: After establishing the design domain, the maximum material usage β of the power distribution network is set.0 Β-cyclodextrin for each step in the course of growth(k) The conductivity of the base structure and the power distribution network is set to σ, respectively.0 And σp Setting initial value X of design variable0 Minimum XMin Maximum XMax . 5.2.2) Growth competition: growth competition is the process of growing competition, i.e. the growth of the flexible growth unit to obtain optimal geometric parameters; the local optimization is a process of optimizing the width of all the flexible growth units; and the global optimization is the process of optimizing the width of all the flexible growth units. 5.2.3) Local reconstruction: configuration reconstruction takes place in the growing process, is used for deciding whether the terminal of certain flexible growth unit continues to grow out the next stage unit. Setting two thresholds: growth threshold Wb And degradation threshold Wd : Newly grown unit intermediate width value t2 ＞ Wb , Then the newly grown cell will be retained; when t2 ＜ Wd , Newly grown cells will be removed; when W is usedd ≤ t2 ≤ Wb The newly grown unit will be retained, but the end thereof cannot grow to the next stage unit anymore. Growth race and configuration reconstruction continues iterate with MMA algorithm updates until the amount of material reaches the maximum β set at initialization.0 , An optimized design of an airborne computer board-level low-direct-current impedance coplanar electromagnetic bandgap power distribution network formed by the growth of the flexible growth unit is obtained.