Power device structure design method and system based on response surface and NSGA-Ⅲ algorithm

Through the combination of the response surface and the NSGA-III algorithm, the structural design of power devices is optimized, and the problems of low power device design efficiency and poor multi-objective optimization effects in the existing technology are solved, and an efficient and simplified design process is achieved, which is suitable for multiple and multi-purpose optimization. Target optimization.

CN114912407BActive Publication Date: 2024-11-05GUILIN UNIV OF ELECTRONIC TECH +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210424443.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-22
Publication Date
2024-11-05
Estimated Expiration
2042-04-22

AI Technical Summary

Technical Problem

In the prior art, the optimization efficiency of power device structure design is low, the simulation method is complex and not suitable for multi-objective optimization, it cannot effectively reflect details, and the mathematical analytical methods are low in versatility, making it difficult to meet the multi-objective optimization needs.

Method used

Using the design method based on the response surface and NSGA-III algorithm, the packaging form and chip layout information of the power device are obtained, the design variables and fixed parameters are determined, the design mathematical model is fitted using the response surface method, and the NSGA-III algorithm is used to solve it. Optimize the power device structure, combine finite element simulation and mathematical analysis methods to achieve multi-objective optimization.

Benefits of technology

It greatly improves design efficiency, shortens the design cycle, reduces professional threshold, is suitable for different models, can effectively reflect the details of power device structure optimization, and is suitable for high-dimensional target optimization problems, meeting multi-objective optimization of 2 to 15 goals need.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114912407B_ABST
    Figure CN114912407B_ABST
Patent Text Reader

Abstract

The present invention provides a power device structure design method and system based on response surface and NSGA-Ⅲ algorithm, wherein the power device structure design method based on response surface and NSGA-Ⅲ algorithm includes: obtaining the packaging form, packaging structure and chip layout information of the power device; determining the size information of the packaging structure and the chip layout; obtaining the design goal; determining the design variables and fixed parameters in the packaging structure and the chip layout size information according to the design goal; according to the design variables and fixed parameters, the response surface method is used to fit the design mathematical model; according to the design mathematical model, the NSGA-Ⅲ algorithm is used to calculate the design mathematical model to obtain the solution set of the design mathematical model, so as to design the power device structure according to the solution set. This application combines finite element simulation and mathematical analytical methods to meet multi-objective optimization while avoiding blind large-scale simulation, without the need to derive the target expression according to the subject theory and model structure, and can provide guidance for the design of other electronic packaging devices.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the fields of electronic packaging and computer technology, and in particular to a power device structure design method and system based on a response surface and NSGA-III algorithm. Background Art

[0002] In the prior art, if a simple simulation optimization method is used for the structural design optimization of power devices, a large amount of simulation work needs to be performed, and the efficiency of relying solely on simulation is low, which leads to a long design cycle; if a simple mathematical analysis method is used, it is necessary to derive the optimization target expression based on the subject theory and model structure, the optimization process is complicated, and requires strong professional capabilities. At the same time, the simple mathematical analysis method is not suitable for different models, has low versatility, and cannot reflect the details of the power device structural optimization design. In addition, the structural optimization method for power devices in the prior art does not have an ideal optimization effect for multiple optimization targets.

[0003] Therefore, how to propose a solution suitable for multi-objective optimization and improve the efficiency of power device packaging structure optimization design has become a problem that needs to be solved urgently. Summary of the invention

[0004] In order to solve the above technical problems, a first aspect of the present invention proposes a power device structure design method based on response surface and NSGA-III algorithm.

[0005] The second aspect of the present invention also proposes a power device structure design system based on response surface and NSGA-III algorithm.

[0006] The third aspect of the present invention also proposes a power device structure design system based on response surface and NSGA-III algorithm.

[0007] A fourth aspect of the present invention also provides a readable storage medium.

[0008] In view of this, the first aspect of the present invention proposes a power device structure design method based on a response surface and NSGA-III algorithm, including: obtaining a packaging form, a packaging structure and a chip layout information of a power device; determining the size information of the packaging structure and the size information of the chip layout; obtaining a design goal; determining the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design goal; using the response surface method to fit a design mathematical model of the design goal according to the design variables and the fixed parameters; using the NSGA-III algorithm to calculate the design mathematical model according to the design mathematical model to obtain a solution set of the design mathematical model, so as to design the power device structure according to the solution set.

[0009] According to the power device structure design method based on response surface and NSGA-III algorithm provided by the present invention, the packaging form, packaging structure and chip layout information of the power device are obtained, and then the size information of the packaging structure and the chip layout can be determined according to the above information, so that the power device can be modeled in the simulation software according to the various size information, so as to facilitate subsequent simulation analysis; then the design target and performance are obtained to determine which performance of the power device needs to be optimized, such as the maximum stress of the chip, the maximum temperature of the chip and / or the module inductance; then according to the determined design target, the size parameters of the packaging structure and the chip layout are determined, and the parameters with greater influence on the design target are selected as design variables, and the parameters with less influence are selected as fixed parameters, and then according to the design variables and fixed parameters, the variable experimental combination is determined based on the response surface method such as Box-Behnken (BBD) and central composite design (CCD) for calculation to obtain a design mathematical model corresponding to the design target, and then the design mathematical model is solved by the NSGA-III algorithm to obtain the Pareto optimal solution set of the design mathematical model, and then the optimal size parameters can be selected according to actual conditions to perform structural optimization design of the power device. This application combines the finite element simulation method and the mathematical analysis method. Compared with the simple simulation optimization method, the efficiency is greatly improved, the blind large-scale simulation is avoided, and the design cycle is shortened; compared with the simple mathematical analysis method, there is no need to derive the optimization target expression according to the subject theory and model structure, which greatly simplifies the optimization process, reduces the professional threshold of the user, is applicable to different models, has good versatility, and can better reflect the details of the power device structure optimization design. The special feature of the present invention is that it combines the response surface method and the new generation NSGA-Ⅲ algorithm to optimize the structure of the power device and applies it to the power packaging field. The feature is that it combines the finite element simulation method and the mathematical analysis method; the optimization target expression can be directly obtained by the software without the need to derive it according to the subject theory and model structure; it is also suitable for solving high-dimensional optimization target problems. The present invention combines the finite element simulation method and the mathematical analysis method. Compared with the simple simulation optimization method, the efficiency is greatly improved, the blind large-scale simulation is avoided, and the design cycle is shortened; compared with the simple mathematical analysis method, there is no need to derive the optimization target expression according to the subject theory and model structure, which greatly simplifies the optimization process, reduces the professional threshold of the user, is applicable to different models, has good versatility, and can better reflect the details of the power device structure optimization design. In addition, the power device structure design method based on the response surface and NSGA-III algorithm proposed in the present invention has better optimization effect on high-dimensional target optimization (i.e., three or more optimization targets). The present invention can solve the problem of unsatisfactory optimization effect for high-dimensional target optimization problems in the prior art, can meet the multi-objective optimization requirements with 2 to 15 targets, and can achieve satisfactory results.

[0010] Among them, Pareto Optimality, also known as Pareto efficiency, refers to an ideal state of resource allocation. It assumes that a group of people and allocable resources change from one allocation state to another, without making anyone's situation worse, making at least one person better off. This is Pareto improvement or Pareto optimization.

[0011] The power device structure design method based on response surface and NSGA-III algorithm proposed in this application is not limited to the structural optimization of power devices. This method can also provide guidance for the optimization design of other electronic packaging devices.

[0012] In addition, the power device structure design method based on the response surface and NSGA-III algorithm in the above technical solution provided by the present invention may also have the following additional technical features:

[0013] In the above technical solution, the steps of obtaining the packaging form, packaging structure and chip layout information of the power device specifically include: obtaining the packaging form of the actual power device to be structurally designed; obtaining the packaging structure of the actual power device, and the dimensions of each layer of the packaging structure; obtaining the number of chips of the actual power device, the chip position and the distance between adjacent chips.

[0014] In this technical solution, the packaging form, packaging structure and chip layout information of the power device are obtained. The packaging form of the power device includes but is not limited to single-sided packaging, double-sided packaging, and stacked packaging. Other packaging forms are also applicable to the structural design method in this application. The packaging structure of the power device includes but is not limited to a base plate, a ceramic substrate, a chip, a buffer layer, and a solder layer. The specific packaging structure is determined according to the actual power device to be structurally designed. For example, the base plate generally only exists in a single-sided package, and the buffer layer generally only exists in the heat dissipation structure of a double-sided package. In addition, the solder layer can be used to connect the base plate to the ceramic substrate, the ceramic substrate to the chip, the chip to the buffer layer, and the buffer layer to the ceramic substrate. The chip layout includes the number of chips, the chip position, and the distance between adjacent chips. The above dimensions are obtained to facilitate modeling of the power device based on the above information for subsequent simulation.

[0015] In the above technical solution, the design target includes at least one or more of the maximum stress of the chip, the maximum temperature of the chip, and the module inductance.

[0016] In this technical solution, design targets are obtained, including but not limited to one or more of chip maximum stress, chip maximum temperature, and module inductance. After the design targets are determined, the design variables, fixed parameters, and performance can be determined to optimize the connection structure of the power device based on the design targets.

[0017] In any of the above technical solutions, before the step of determining the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design objectives, it also includes: determining the variable size parameters in the size information of the packaging structure and the size information of the chip layout; determining the numerical variation range of the variable size parameters.

[0018] In this technical solution, by determining the variable size parameters in the size information of the packaging structure and the size information of the chip layout, and then determining the numerical variation range of the variable size parameters, the degree of influence of the variable parameter size on the design target can be determined based on the numerical variation range of the variable size parameters, so as to determine the design variables and fixed parameters based on the degree of influence.

[0019] In the above technical solution, the step of determining the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design objectives specifically includes: after determining the numerical variation range of the variable size parameters, according to the degree of influence of the variable size parameters on the design objectives, obtaining the size parameters as design variables and the size parameters as fixed parameters.

[0020] In this technical solution, by determining the degree of influence of variable size parameters on the design objectives, the design variables and fixed parameters are determined to facilitate the subsequent determination of variable experimental combinations, thereby reducing the workload of simulation and experiments. There is no need to rely on mathematical theories for a large number of calculations, saving human resources and making the design process of the power device structure simpler.

[0021] In any of the above technical solutions, the degree of influence of the variable size parameter on the design target is obtained through univariate sensitivity analysis and / or artificial experience; when the parameter values ​​of the design variable and the fixed parameter undergo the same changes, the degree of influence of the design variable on the design target is greater than the degree of influence of the fixed parameter on the design target.

[0022] In this technical solution, after determining the numerical variation range of the variable dimensional parameters, the dimensional parameters that have a greater impact on the optimization design target are screened out as optimization design variables based on design experience or through sensitivity analysis and other methods, and the dimensional parameters that have a smaller impact on the target are selected as fixed design parameters, so as to facilitate the calculation of the design mathematical model of the optimization design target based on the design variables and fixed parameters.

[0023] In any of the above technical solutions, the steps of fitting the design mathematical model of the design target by the response surface method according to the design variables and fixed parameters specifically include: determining the variable experimental combination by the response surface method according to the numerical variation range of the design target, the design variables and the variable size parameters; performing finite element simulation according to the variable experimental combination to obtain simulation results; calculating the regression coefficient by the response surface method according to the simulation results; obtaining the objective function and constraint function of the design target according to the regression coefficient; and obtaining the design mathematical model according to the objective function and the constraint function.

[0024] In this technical solution, by determining the design objectives, the design variables and fixed parameters are determined, and then the response surface methods such as Box-Behnken (BBD) and central composite design (CCD) can be used to determine the variable experimental combination and obtain the response value results through simulation, and then the response surface method is used for regression analysis to obtain the regression coefficient, so as to obtain the objective function and constraint function, and then a multi-objective optimization mathematical model of the power device packaging structure, that is, the design mathematical model, can be formed according to the objective function and constraint function.

[0025] Among them, response surface analysis, also known as response surface design method (RSM), is a statistical method that uses reasonable experimental design methods to obtain certain data through experiments, uses multivariate quadratic regression equations to fit the functional relationship between factors and response values, and seeks optimal process parameters through analysis of regression equations to solve multivariate problems.

[0026] The second aspect of the present invention provides a power device structure design system based on response surface and NSGA-III algorithm, including: an information acquisition module, used to obtain the packaging form, packaging structure and chip layout information of the power device, determine the size information of the packaging structure and the size information of the chip layout, obtain the design goal, and determine the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design goal; a calculation module, used to use the response surface method to fit the design mathematical model of the design goal according to the design variables and fixed parameters; according to the design mathematical model, use the NSGA-III algorithm to calculate the design mathematical model to obtain the solution set of the design mathematical model, so as to design the power device structure according to the solution set.

[0027] The power device structure design system based on the response surface and NSGA-Ⅲ algorithm provided by the technical solution of the present invention includes: an information acquisition module and a calculation module. Specifically, the information acquisition module is used to obtain the packaging form, packaging structure and chip layout information of the power device, determine the size information of the packaging structure and the size information of the chip layout, obtain the design target, and determine the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design target; the calculation module is used to fit the design mathematical model of the design target using the response surface method according to the design variables and fixed parameters; according to the design mathematical model, the design mathematical model is calculated using the NSGA-Ⅲ algorithm to obtain the solution set of the design mathematical model, so as to design the power device structure according to the solution set. At the same time, the power device structure design system based on the response surface and NSGA-Ⅲ algorithm provided by the technical solution of the present invention, because it is used to implement the steps of the power device structure design method based on the response surface and NSGA-Ⅲ algorithm provided by the first aspect of the present invention, the power device structure design system based on the response surface and NSGA-Ⅲ algorithm has all the technical effects of the power device structure design method based on the response surface and NSGA-Ⅲ algorithm, which will not be repeated here.

[0028] In any of the above technical solutions, the information acquisition module is specifically used to: obtain the packaging form of the actual power device to be structurally designed; obtain the packaging structure of the actual power device, and the dimensions of each layer of the packaging structure; obtain the number of chips of the actual power device, the chip position and the distance between adjacent chips.

[0029] In this technical solution, the packaging form, packaging structure and chip layout information of the power device are obtained. The packaging form of the power device includes but is not limited to single-sided packaging, double-sided packaging, and stacked packaging. Other packaging forms are also applicable to the structural design method in this application. The packaging structure of the power device includes but is not limited to a base plate, a ceramic substrate, a chip, a buffer layer, and a solder layer. The specific packaging structure is determined according to the actual power device to be structurally designed. For example, the base plate generally only exists in a single-sided package, and the buffer layer generally only exists in the heat dissipation structure of a double-sided package. In addition, the solder layer can be used to connect the base plate to the ceramic substrate, the ceramic substrate to the chip, the chip to the buffer layer, and the buffer layer to the ceramic substrate. The chip layout includes the number of chips, the chip position, and the distance between adjacent chips. The above dimensions are obtained to facilitate modeling of the power device based on the above information for subsequent simulation.

[0030] In any of the above technical solutions, the design objectives include at least one or more of the maximum stress of the chip, the maximum temperature of the chip, and the module inductance.

[0031] In this technical solution, design targets are obtained, including but not limited to one or more of chip maximum stress, chip maximum temperature, and module inductance. After the design targets are determined, the design variables, fixed parameters, and performance can be determined to optimize the connection structure of the power device based on the design targets.

[0032] In any of the above technical solutions, the information acquisition module is also used to: determine the variable size parameters in the size information of the packaging structure and the size information of the chip layout; and determine the value variation range of the variable size parameters.

[0033] In this technical solution, by determining the variable size parameters in the size information of the packaging structure and the size information of the chip layout, and then determining the numerical variation range of the variable size parameters, it is possible to determine the degree of influence of the variable parameter size on the design target based on the numerical variation range of the variable size parameters, so as to determine the design variables and fixed parameters based on the degree of influence. In any of the above technical solutions, the information acquisition module is also specifically used to: after determining the numerical variation range of the variable size parameters, obtain the size parameters as design variables and the size parameters as fixed parameters based on the degree of influence of the variable size parameters on the design target.

[0034] In this technical solution, by determining the degree of influence of variable size parameters on the design objectives, the design variables and fixed parameters are determined to facilitate the subsequent determination of variable experimental combinations, thereby reducing the workload of simulation and experiments without relying on mathematical theories for a large number of calculations, saving human resources and making the design process of the power device structure simpler.

[0035] In any of the above technical solutions, the degree of influence of the variable size parameter on the design target is obtained through univariate sensitivity analysis and / or artificial experience; when the parameter values ​​of the design variable and the fixed parameter undergo the same changes, the degree of influence of the design variable on the design target is greater than the degree of influence of the fixed parameter on the design target.

[0036] In this technical solution, after determining the numerical variation range of the variable dimensional parameters, the dimensional parameters that have a greater impact on the optimization design target are screened out as optimization design variables based on design experience or through sensitivity analysis and other methods, and the dimensional parameters that have a smaller impact on the target are selected as fixed design parameters, so as to facilitate the calculation of the design mathematical model of the optimization design target based on the design variables and fixed parameters.

[0037] In any of the above technical solutions, the calculation module is specifically used for: determining the variable experimental combination by using the response surface method according to the numerical variation range of the design objectives, design variables and variable size parameters; performing finite element simulation according to the variable experimental combination to obtain simulation results; calculating the regression coefficient by using the response surface method according to the simulation results; obtaining the objective function and constraint function of the design objective according to the regression coefficient; and obtaining the design mathematical model according to the objective function and constraint function.

[0038] In this technical solution, by determining the design objectives, the design variables and fixed parameters are determined, and then the response surface methods such as Box-Behnken (BBD) and central composite design (CCD) can be used to determine the variable experimental combination and obtain the response value results through simulation, and then the response surface method is used for regression analysis to obtain the regression coefficient, so as to obtain the objective function and constraint function, and then a multi-objective optimization mathematical model of the power device packaging structure, that is, the design mathematical model, can be formed according to the objective function and constraint function.

[0039] Among them, response surface analysis, also known as response surface design method (RSM), is a statistical method that uses reasonable experimental design methods to obtain certain data through experiments, uses multivariate quadratic regression equations to fit the functional relationship between factors and response values, and seeks optimal process parameters through analysis of regression equations to solve multivariate problems.

[0040] The third aspect of the present invention provides a power device structure design system based on response surface and NSGA-III algorithm, comprising: a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps specified in the power device structure design method based on response surface and NSGA-III algorithm of any of the above-mentioned technical solutions are implemented.

[0041] The power device structure design system based on response surface and NSGA-Ⅲ algorithm provided by the technical solution of the present invention includes a memory and a processor and a program stored in the memory and executable on the processor, and when the program is executed by the processor, the steps defined by any of the above-mentioned power device structure design methods based on response surface and NSGA-Ⅲ algorithm are implemented. At the same time, since the power device structure design system based on response surface and NSGA-Ⅲ algorithm of the present application can implement the steps defined by any of the above-mentioned power device structure design methods based on response surface and NSGA-Ⅲ algorithm, the power device structure design system based on response surface and NSGA-Ⅲ algorithm provided by the technical solution has all the beneficial effects of the power device structure design method based on response surface and NSGA-Ⅲ algorithm provided in any of the above-mentioned technical solutions.

[0042] The fourth aspect of the present invention provides a readable storage medium having programs and / or instructions stored thereon, which, when executed by a processor, implement the steps of the power device structure design method based on the response surface and NSGA-III algorithm in any of the above-mentioned technical solutions.

[0043] The readable storage medium provided according to the technical solution of the present invention can implement the steps of the power device structure design method based on response surface and NSGA-III algorithm in any of the above-mentioned technical solutions when the program and / or instructions stored thereon are executed by the processor. Therefore, it has all the beneficial technical effects of the power device structure design method based on response surface and NSGA-III algorithm, which will not be repeated here.

[0044] Additional aspects and advantages of the present invention will become apparent from the following description or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0046] Figure 1 is a flow chart of a power device structure design method based on a response surface and NSGA-III algorithm according to an embodiment of the present invention;

[0047] Figure 2 is a block diagram of a power device structure design system based on a response surface and NSGA-III algorithm according to an embodiment of the present invention;

[0048] Figure 3 is a block diagram of a power device structure design system based on a response surface and NSGA-III algorithm according to an embodiment of the present invention;

[0049] Figure 4 is a flow chart of a power device structure design method based on a response surface and NSGA-III algorithm according to another embodiment of the present invention;

[0050] Figure 5 is a schematic diagram of a central combination design model according to another embodiment of the present invention;

[0051] Figure 6 is a vertical structural diagram of a power device according to another embodiment of the present invention;

[0052] Figure 7 is a simplified chip layout diagram of a power device according to another embodiment of the present invention;

[0053] Figure 8 is a schematic diagram of a solution set of an objective function according to another embodiment of the present invention;

[0054] Fig. 9 is a distribution diagram of the values ​​of each design variable in the solution set of the objective function according to another embodiment of the present invention;

[0055] Fig.10 is a comparison diagram of power device performance before and after optimization according to another embodiment of the present invention;

[0056] Fig.11 It is a flowchart of the NSGA-III algorithm according to another embodiment of the present invention.

[0057] in, Figure 2 and Figure 3 , Figure 5 and Figure 6 The corresponding relationship between the reference numerals and the component names is as follows:

[0058] 200 Power device structure design system based on response surface and NSGA-Ⅲ algorithm, 202 Information acquisition module, 204 Calculation module, 300 Power device structure design system based on response surface and NSGA-Ⅲ algorithm, 302 Memory, 304 Processor, 602 Copper layer, 604 IGBT chip, 606 Solder layer 1, 608 Solder layer 2, 610 Solder layer 3, 612 Ceramic layer, 614 Diode chip, 616 Buffer layer. DETAILED DESCRIPTION

[0059] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0060] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0061] Refer to the following Figures 1 to 11 The present invention describes a power device structure design method and system based on a response surface and NSGA-III algorithm in some embodiments of the present invention.

[0062] In view of this, the first embodiment of the present invention proposes a power device structure design method based on response surface and NSGA-III algorithm, such as Figure 1 As shown, including:

[0063] S102, obtaining packaging form, packaging structure and chip layout information of the power device;

[0064] S104, determining the size information of the package structure and the size information of the chip layout;

[0065] S106, obtaining a design target;

[0066] S108, determining design variables and fixed parameters in the size information of the package structure and the size information of the chip layout according to the design goal;

[0067] S110, according to the design variables and fixed parameters, a design mathematical model of the design target is obtained by fitting using the response surface method;

[0068] S112, according to the design mathematical model, using the NSGA-III algorithm to calculate the design mathematical model to obtain a solution set of the design mathematical model, so as to design a power device structure according to the solution set.

[0069] According to the power device structure design method based on the response surface and NSGA-III algorithm provided in this embodiment, by obtaining the packaging form, packaging structure, and chip layout information of the power device, the size information of the packaging structure and the chip layout can be determined according to the above information, so that the power device can be modeled in the simulation software according to the various size information, so as to facilitate subsequent simulation analysis; then the design goal and performance are obtained to determine which performance of the power device needs to be optimized, such as the maximum stress of the chip, the maximum temperature of the chip and / or the module inductance; then according to the determined design goal, the size parameters of the packaging structure and the chip layout are determined, and the parameters with greater influence on the design goal are selected as design variables, and the parameters with less influence are selected as fixed parameters, and then according to the design variables and fixed parameters, the variable experimental combination is determined based on the response surface method such as Box-Behnken (BBD) and central composite design (CCD) for calculation to obtain a design mathematical model corresponding to the design goal, and then the design mathematical model is solved by the NSGA-III algorithm to obtain the Pareto optimal solution set of the design mathematical model, and then the optimal size parameters can be selected according to actual conditions to perform structural optimization design of the power device. This application combines finite element simulation method and mathematical analysis method, which greatly improves efficiency compared with pure simulation optimization method, avoids blind large-scale simulation, and shortens the design cycle; compared with pure mathematical analysis method, there is no need to derive the optimization target expression according to subject theory and model structure, which greatly simplifies the optimization process, reduces the professional threshold of users, is applicable to different models, has good versatility, and can better reflect the details of power device structure optimization design. The special feature of the present invention is that it combines the response surface method and the new generation NSGA-Ⅲ algorithm to optimize the structure of power devices and applies it to the field of power packaging. The characteristics are that it combines finite element simulation and mathematical analysis methods; the optimization target expression can be directly obtained by the software without derivation according to subject theory and model structure; it is also suitable for solving high-dimensional optimization target problems. The method and method characteristics used in the present invention should be highlighted. The present invention combines the finite element simulation method and the mathematical analysis method. Compared with the simple simulation optimization method, the efficiency is greatly improved, the blind large-scale simulation is avoided, and the design cycle is shortened; compared with the simple mathematical analysis method, there is no need to derive the optimization target expression according to the subject theory and model structure, which greatly simplifies the optimization process, reduces the professional threshold of the user, is applicable to different models, has good versatility, and can better reflect the details of the power device structure optimization design. In addition, the power device structure design method based on the response surface and NSGA-Ⅲ algorithm proposed in the present invention has a better optimization effect on high-dimensional target optimization (i.e., three or more optimization targets). The present invention can solve the problem of unsatisfactory optimization effect in the face of high-dimensional target optimization problems in the prior art, can meet the multi-objective optimization needs with 2 to 15 targets, and can achieve satisfactory results.

[0070] Among them, Pareto Optimality, also known as Pareto efficiency, refers to an ideal state of resource allocation. It assumes that a group of people and allocable resources change from one allocation state to another, without making anyone's situation worse, making at least one person better off. This is Pareto improvement or Pareto optimization.

[0071] The power device structure design method based on response surface and NSGA-III algorithm proposed in this application is not limited to the structural optimization of power devices. This method can also provide guidance for the optimization design of other electronic packaging devices.

[0072] In the above embodiment, the step of obtaining the packaging form, packaging structure and chip layout information of the power device specifically includes: obtaining the packaging form of the actual power device to be structurally designed; obtaining the packaging structure of the actual power device and the dimensions of each layer of the packaging structure; obtaining the number of chips, chip position and distance between adjacent chips of the actual power device. In this embodiment, the packaging form, packaging structure and chip layout information of the power device are obtained. The packaging form of the power device includes but is not limited to single-sided packaging, double-sided packaging, and stacked packaging. Other packaging forms are also applicable to the structural design method in this application. The packaging structure of the power device includes but is not limited to a base plate, a ceramic substrate, a chip, a buffer layer, and a welding layer. The specific packaging structure is determined according to the actual power device to be structurally designed. For example, the base plate generally only exists in a single-sided package, and the buffer layer generally only exists in the heat dissipation structure of the double-sided package. In addition, the welding layer can be used to connect the base plate to the ceramic substrate, the ceramic substrate to the chip, the chip to the buffer layer, and the buffer layer to the ceramic substrate. The chip layout includes the number of chips, the chip position and the distance between adjacent chips. The above dimensions are obtained to facilitate modeling of the power device according to the above information for subsequent simulation.

[0073] In the above embodiment, the design target includes at least one or more of the maximum stress of the chip, the maximum temperature of the chip, and the module inductance.

[0074] In this embodiment, the design goals are obtained, and the design goals include but are not limited to one or more of the maximum stress of the chip, the maximum temperature of the chip, and the module inductance. After the design goals are determined, the design variables, fixed parameters and performance can be determined to optimize the connection structure of the power device based on the design goals.

[0075] In any of the above technical solutions, before the step of determining the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design objectives, it also includes: determining the variable size parameters in the size information of the packaging structure and the size information of the chip layout; determining the numerical variation range of the variable size parameters.

[0076] In this technical solution, by determining the variable size parameters in the size information of the packaging structure and the size information of the chip layout, and then determining the numerical variation range of the variable size parameters, the degree of influence of the variable parameter size on the design target can be determined based on the numerical variation range of the variable size parameters, so as to determine the design variables and fixed parameters based on the degree of influence.

[0077] In the above embodiment, the step of determining the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design objectives specifically includes: after determining the numerical variation range of the variable size parameters, according to the degree of influence of the variable size parameters on the design objectives, obtaining the size parameters as design variables and the size parameters as fixed parameters.

[0078] In this embodiment, by determining the degree of influence of variable size parameters on the design objectives, the design variables and fixed parameters are determined to facilitate the subsequent determination of variable experimental combinations, thereby reducing the workload of simulation and experiments. There is no need to rely on mathematical theories for a large number of calculations, saving human resources and making the design process of the power device structure simpler.

[0079] In any of the above embodiments, the degree of influence of the variable size parameter on the design objective is obtained through univariate sensitivity analysis and / or manually based on experience; when the parameter values ​​of the design variable and the fixed parameter undergo the same changes, the degree of influence of the design variable on the design objective is greater than the degree of influence of the fixed parameter on the design objective.

[0080] In this embodiment, after determining the numerical variation range of the variable dimensional parameters, the dimensional parameters that have a greater impact on the optimization design target are screened out as optimization design variables based on design experience or through methods such as sensitivity analysis, and the dimensional parameters that have a smaller impact on the target are selected as fixed design parameters. This facilitates the calculation of the design mathematical model for the optimization design target based on the design variables and fixed parameters.

[0081] In the above embodiment, the steps of fitting the design mathematical model of the design target by the response surface method according to the design variables and fixed parameters specifically include: determining the variable experimental combination by the response surface method according to the numerical variation range of the design target, the design variables and the variable size parameters; performing finite element simulation according to the variable experimental combination to obtain simulation results; calculating the regression coefficient by the response surface method according to the simulation results; obtaining the objective function and constraint function of the design target according to the regression coefficient; and obtaining the design mathematical model according to the objective function and the constraint function.

[0082] In this embodiment, by determining the design objectives, the design variables and fixed parameters are determined, and then the response surface methods such as Box-Behnken (BBD) and central composite design (CCD) can be used to determine the variable experimental combination and obtain the response value results through simulation, and then the response surface method is used to perform regression analysis to obtain the regression coefficient, thereby obtaining the objective function and constraint function, and then a multi-objective optimization mathematical model of the power device packaging structure, that is, a design mathematical model, can be formed according to the objective function and constraint function.

[0083] Among them, response surface analysis, also known as response surface design method (RSM), is a statistical method that uses reasonable experimental design methods to obtain certain data through experiments, uses multivariate quadratic regression equations to fit the functional relationship between factors and response values, and seeks optimal process parameters through analysis of regression equations to solve multivariate problems.

[0084] The second embodiment of the present invention provides a power device structure design system 200 based on response surface and NSGA-III algorithm, such as Figure 2 As shown, it includes: an information acquisition module 202, which is used to obtain the packaging form, packaging structure and chip layout information of the power device, determine the size information of the packaging structure and the size information of the chip layout, obtain the design target, and determine the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design target; a calculation module 204, which is used to fit the design mathematical model of the design target using the response surface method according to the design variables and fixed parameters; according to the design mathematical model, the NSGA-Ⅲ algorithm is used to calculate the design mathematical model to obtain the solution set of the design mathematical model, so as to design the power device structure according to the solution set.

[0085] The power device structure design system based on the response surface and NSGA-Ⅲ algorithm provided in accordance with the embodiment of the present invention comprises: an information acquisition module 202 and a calculation module 204. Specifically, the information acquisition module 202 is used to obtain the packaging form, packaging structure and chip layout information of the power device, determine the size information of the packaging structure and the size information of the chip layout, obtain the design target, and determine the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design target; the calculation module 204 is used to fit the design mathematical model of the design target using the response surface method according to the design variables and fixed parameters; according to the design mathematical model, the design mathematical model is calculated using the NSGA-Ⅲ algorithm to obtain the solution set of the design mathematical model, so as to design the power device structure according to the solution set. At the same time, the power device structure design system based on the response surface and NSGA-Ⅲ algorithm provided in accordance with the embodiment of the present invention, because it is used to implement the steps of the power device structure design method based on the response surface and NSGA-Ⅲ algorithm provided in the embodiment of the first aspect of the present invention, the power device structure design system based on the response surface and NSGA-Ⅲ algorithm has all the technical effects of the power device structure design method based on the response surface and NSGA-Ⅲ algorithm, which will not be repeated here.

[0086] In any of the above embodiments, the information acquisition module 202 is specifically used to: obtain the packaging form of the actual power device to be structurally designed; obtain the packaging structure of the actual power device, and the dimensions of each layer of the packaging structure; obtain the number of chips of the actual power device, the chip position and the distance between adjacent chips.

[0087] In this embodiment, the packaging form, packaging structure and chip layout information of the power device are obtained. The packaging form of the power device includes but is not limited to single-sided packaging, double-sided packaging, and stacked packaging. Other packaging forms are also applicable to the structural design method in this application. The packaging structure of the power device includes but is not limited to a base plate, a ceramic substrate, a chip, a buffer layer, and a solder layer. The specific packaging structure is determined according to the actual power device to be connected to the structural design. For example, the base plate generally only exists in a single-sided package, and the buffer layer generally only exists in the heat dissipation structure of the double-sided package. In addition, the solder layer can be used to connect the base plate to the ceramic substrate, the ceramic substrate to the chip, the chip to the buffer layer, and the buffer layer to the ceramic substrate. The chip layout includes the number of chips, the chip position and the distance between adjacent chips. The above dimensions are obtained to facilitate modeling of the power device according to the above information for subsequent simulation.

[0088] In any of the above embodiments, the design target includes at least one or more of the maximum stress of the chip, the maximum temperature of the chip, and the module inductance.

[0089] In this embodiment, the design goals are obtained, and the design goals include but are not limited to one or more of the maximum stress of the chip, the maximum temperature of the chip, and the module inductance. After the design goals are determined, the design variables, fixed parameters and performance can be determined to optimize the connection structure of the power device based on the design goals.

[0090] In any of the above technical solutions, the information acquisition module is also used to: determine the variable size parameters in the size information of the packaging structure and the size information of the chip layout; and determine the value variation range of the variable size parameters.

[0091] In this technical solution, by determining the variable size parameters in the size information of the package structure and the size information of the chip layout, and then determining the value variation range of the variable size parameters, the influence of the variable parameter size on the design target can be determined according to the value variation range of the variable size parameters, so as to determine the design variables and fixed parameters according to the influence degree. In any of the above embodiments, the information acquisition module 202 is also specifically used to: after determining the value variation range of the variable size parameters, according to the influence degree of the variable size parameters on the design target, obtain the size parameters as design variables and the size parameters as fixed parameters.

[0092] In this embodiment, by determining the degree of influence of variable size parameters on the design objectives, the design variables and fixed parameters are determined to facilitate the subsequent determination of variable experimental combinations, thereby reducing the workload of simulation and experiments without relying on mathematical theories for a large number of calculations, saving human resources and making the design process of the power device structure simpler.

[0093] In the above embodiment, the influence degree of the variable size parameter on the design target is obtained through univariate sensitivity analysis and / or manually based on experience; when the parameter values ​​of the design variable and the fixed parameter undergo the same changes, the influence degree of the design variable on the design target is greater than the influence degree of the fixed parameter on the design target.

[0094] In this embodiment, after determining the numerical variation range of the variable dimensional parameters, the dimensional parameters that have a greater impact on the optimization design target are screened out as optimization design variables based on design experience or through methods such as sensitivity analysis, and the dimensional parameters that have a smaller impact on the target are selected as fixed design parameters. This facilitates the calculation of the design mathematical model for the optimization design target based on the design variables and fixed parameters.

[0095] In any of the above embodiments, the calculation module 204 is specifically used to: determine the variable experimental combination using the response surface method according to the numerical variation range of the design objectives, design variables, and variable dimensional parameters; perform finite element simulation according to the variable experimental combination to obtain simulation results; calculate the regression coefficient using the response surface method according to the simulation results; obtain the objective function and constraint function of the design objective according to the regression coefficient; and obtain the design mathematical model according to the objective function and the constraint function.

[0096] In this embodiment, by determining the design objectives, the design variables and fixed parameters are determined, and then the response surface methods such as Box-Behnken (BBD) and central composite design (CCD) can be used to determine the variable experimental combination and obtain the response value results through simulation, and then the response surface method is used to perform regression analysis to obtain the regression coefficient, thereby obtaining the objective function and constraint function, and then a multi-objective optimization mathematical model of the power device packaging structure, that is, a design mathematical model, can be formed according to the objective function and constraint function.

[0097] Among them, response surface analysis, also known as response surface design method (RSM), is a statistical method that uses reasonable experimental design methods to obtain certain data through experiments, uses multivariate quadratic regression equations to fit the functional relationship between factors and response values, and seeks optimal process parameters through analysis of regression equations to solve multivariate problems.

[0098] The third aspect of the present invention provides a power device structure design system 300 based on response surface and NSGA-III algorithm, such as Figure 3 As shown, it includes: a memory 302, a processor 304, and a program stored in the memory 302 and executable on the processor 304. When the program is executed by the processor 304, the steps defined by the power device structure design method based on the response surface and NSGA-III algorithm of any of the above-mentioned embodiments are implemented.

[0099] According to an embodiment of the present invention, a power device structure design system based on a response surface and NSGA-III algorithm is provided, including a memory and a processor and a program stored in the memory and executable on the processor, and when the program is executed by the processor, the steps defined in any of the above-mentioned power device structure design methods based on a response surface and NSGA-III algorithm are implemented. At the same time, since the power device structure design system based on a response surface and NSGA-III algorithm of the present application can implement the steps defined in any of the above-mentioned power device structure design methods based on a response surface and NSGA-III algorithm, the power device structure design system based on a response surface and NSGA-III algorithm provided in this embodiment has all the beneficial effects of the power device structure design method based on a response surface and NSGA-III algorithm provided in any of the above-mentioned embodiments.

[0100] The fourth aspect of the present invention provides a readable storage medium having programs and / or instructions stored thereon. When the programs and / or instructions are executed by a processor, the steps of the power device structure design method based on the response surface and NSGA-III algorithm in any of the above embodiments are implemented.

[0101] The readable storage medium provided according to an embodiment of the present invention can implement the steps of the power device structure design method based on response surface and NSGA-III algorithm in any of the above-mentioned embodiments when the program and / or instructions stored thereon are executed by a processor. Therefore, it has all the beneficial technical effects of the power device structure design method based on response surface and NSGA-III algorithm, which will not be repeated here.

[0102] The following is a specific example to further introduce the power device structure design method based on the response surface and NSGA-III algorithm provided by the present application.

[0103] This embodiment provides a power device structure optimization method based on response surface and NSGA-III algorithm, such as Figure 4 and Figure 6 As shown, it mainly includes the following steps:

[0104] S402, determine the power device packaging form, packaging structure and chip layout; packaging forms include but are not limited to single-sided packaging and double-sided packaging; packaging structure includes but is not limited to base plate; ceramic substrate, the ceramic substrate includes an upper copper layer, a ceramic layer, and a lower copper layer; chip; buffer layer; solder layer, the solder layer can be used to connect the base plate and the ceramic substrate, the ceramic substrate and the chip, the chip and the buffer layer, and the buffer layer and the ceramic substrate; chip layout includes the number of chips and the approximate location of the chip. Specifically, the power device packaging model can be parameterized in simulation software such as COMSOL, set the heat source and boundary conditions, and continue with subsequent performance analysis.

[0105] S404, determining variable package size parameters and their design space: including the thickness of each layer in the package structure and the distance between chips.

[0106] S406, determining optimization design goals, such as maximum chip stress, maximum chip junction temperature, module inductance, etc.

[0107] S408, based on design experience or through sensitivity analysis and other methods, select the size parameters that have a greater impact on the optimization design target as optimization design variables, and the size parameters that have a smaller impact on the target as fixed design parameters. Specifically, the method of controlling a single variable is as follows: sequentially change a single structural size parameter and fix other parameters, perform simulation through COMSOL and other software, analyze the impact of the size parameter on the optimization design target, select the one with a higher impact as the optimization design variable, and the one with a smaller impact as the fixed design parameter.

[0108] S410, based on the response surface method, the Box-Behnken central composite design model is used to determine the variable experimental combination and obtain the response value results through simulation. The least squares regression analysis is used to obtain the regression coefficient, and the objective function and constraint function are obtained to form a multi-objective optimization mathematical model for the power device packaging structure. Specifically, the Design Expert software is used to select the Box-Behnken central composite design model (BBD), and the high and low values ​​of the optimized design variables and their design dimensions selected in S308 are input to determine the experimental combination of the design variables. The heat source and boundary conditions of the model are set by COMSOL software, and the response value results of each experimental combination are obtained by simulation. The results are input into Design Expert for analysis to obtain the fitting coefficients, and the non-significant influencing factors are removed to obtain the regression equation, i.e., the objective function. The parameter values ​​in the output response list that reflect the rationality of the model and the fitting accuracy should be paid special attention to. The model significance test P-value < 0.05 indicates that the regression effect of the model is significant and the degree of fitting is high; the correction determination coefficient R 2 It reflects the degree to which the response surface fits the given data, R 2 ∈[0,1], and the closer its value is to 1, the better the fitting degree is.

[0109] The Box-Behnken central composite design model (BBD) has treatment combinations located at the midpoint of the edge of the experimental space and has at least three factors, such as Figure 5 The Box-Behnken central composite design model is for three factors, and the points on the graph represent the experimental runs performed.

[0110] S412, using a non-dominated genetic algorithm NSGA-Ⅲ multi-objective genetic algorithm to obtain the Pareto optimal solution set of the multi-objective optimization mathematical model, and then combining it with the actual application conditions, select and determine the optimal package size. Specifically, according to the design space of the design variables and the design target regression equation obtained in step 5, using a non-dominated genetic algorithm NSGA-Ⅲ multi-objective genetic algorithm to obtain the Pareto solution set of the multi-objective optimization mathematical model, and then combining it with the design requirements or actual application conditions, select the optimal solution from the solution set.

[0111] The power device structure design method based on the response surface and NSGA-Ⅲ algorithm provided in this embodiment, through the combination of finite element simulation method and mathematical analysis method, is greatly improved in efficiency compared with the simple simulation optimization method, avoiding blind large-scale simulation and shortening the design cycle; compared with the simple mathematical analysis method, there is no need to derive the optimization target expression according to the subject theory and model structure, which greatly simplifies the optimization process, reduces the professional threshold of users, is applicable to different models, has good versatility, and can better reflect the details of the power device structure optimization design. At present, most multi-objective evolutionary algorithms are more effective in solving problems with low target dimensions, but the proportion of non-dominated solutions will increase exponentially with the increase in the number of targets, resulting in non-dominated solutions occupying most of the positions in the population, greatly slowing down the search process, especially when the number of targets is greater than or equal to 3, that is, facing high-dimensional target optimization problems, the effect becomes unsatisfactory. The main idea of ​​NSGA-Ⅲ is to introduce a reference point mechanism on the basis of NSGA-Ⅱ to retain those non-dominated and close to the reference point population individuals. NSGA-III is suitable for many multi-objective test problems with 2 to 15 targets, and can achieve satisfactory results.

[0112] The power device structure design method based on the response surface and NSGA-III algorithm provided by the present application is further introduced below in conjunction with another specific embodiment.

[0113] This embodiment provides a power device structure optimization method based on response surface and NSGA-III algorithm, which is a power device package structure optimization design method based on response surface and NSGA-III genetic algorithm, and mainly includes the following steps:

[0114] First, determine the power device packaging form, packaging structure and chip layout. This embodiment uses a double-sided heat dissipation half-bridge power module for optimization. Its vertical structure diagram and simplified chip layout diagram are shown in Figure 2. Figure 6 and Figure 7 As shown, two IGBT chips 604 (insulated gate bipolar transistor chips) and two diode chips 614 are included.

[0115] Furthermore, if Figure 6 As shown, the designable size parameters include: the thickness of the copper layer 602 of the ceramic substrate, the thickness of the ceramic layer 612 of the ceramic substrate, the thickness of the solder layer 1606 used to connect the ceramic substrate and the chip, the thickness of the solder layer 2608 used to connect the chip and the buffer layer 616, the thickness of the buffer layer 616, the thickness of the solder layer 3610 used to connect the buffer layer and the ceramic substrate, the distance between the IGBT chip 604 and the diode chip 614, and the distance between the two IGBT chips 604. The design space is shown in Table 1, where the units of each value are mm:

[0116] Table 1 Parameter design space

[0117]

[0118] Furthermore, seven factors, including the thickness X1 of the solder layer 1606 used to connect the ceramic substrate and the chip, the thickness X2 of the solder layer 2608 used to connect the chip and the buffer layer, the thickness X3 of the solder layer 3610 used to connect the buffer layer and the ceramic substrate, the thickness X4 of the buffer layer, the thickness X5 of the ceramic layer of the ceramic substrate, the distance X6 between the IGBT and the diode, and the distance X7 between the two IGBT chips, were selected as design variables, and the thickness of the copper layer of the ceramic substrate was taken as a fixed design parameter.

[0119] Then, the maximum chip stress F and the maximum chip junction temperature T were selected as the optimization design targets. Then, the Box-Behnken central combination design model (BBD) was selected using the DesignExpert software, and the high and low values ​​of the selected optimization design variables and their design dimensions were input to determine the experimental combinations of 62 design variables, as shown in Table 2. The response values ​​of each experimental combination were obtained by simulating the model with COMSOL software and entered into the experimental combination table, as shown in Table 2, where the units of all values ​​without marked units are mm.

[0120] Table 2 Experimental combination table

[0121]

[0122]

[0123]

[0124] By fitting the response value, the model regression coefficient is obtained, and the constant term, cross term, and quadratic term with non-significant influence are removed. Finally, the fitted second-order model regression equation, i.e., the objective function, is obtained:

[0125]

[0126]

[0127] P-value and R of response surface model 2 The values ​​and other relevant fit metrics are shown in Table 3, from which it can be seen that the model is significant, the fit is high, and the accuracy meets the requirements.

[0128] Table 3 Response surface fitting metrics

[0129]

[0130] The NSGA-Ⅲ algorithm is used to find the multi-objective optimization Pareto solution set of the objective function. The population size is set to 150, the maximum evolutionary number is 800, the stop number is 800, the optimal individual coefficient is 0.4, and the fitness function deviation is 1e-8. The Pareto solution set is as follows: Figure 8 shown.

[0131] When selecting the optimal solution of the model, it should be selected according to the actual use conditions. If the focus is on optimizing the mechanical properties of the chip, the solution that minimizes the stress can be selected. If the focus is on the thermal performance of the chip, the solution that minimizes the temperature can be selected. If both performances are taken into account, the solution at the inflection point in the figure can be selected. Table 4 lists the optimal solutions in this optimization set, where the values ​​without marked units are all in mm.

[0132] Table 4 Optimal solution selection

[0133]

[0134] like Fig. 9 As shown, the distribution of the values ​​of each design variable in the optimal solution set is listed. It can be seen that when the ceramic layer is thicker and the buffer layer is thinner, the thermal performance of the chip is better.

[0135] The comparison of the simulation value before optimization and the Pareto solution after optimization in this embodiment, that is, the comparison of the performance of the power device before and after optimization is shown in the figure below: Fig.10 shown.

[0136] The flowchart of the NSGA-Ⅲ algorithm is as follows: Fig.11 As shown, it mainly includes the following steps:

[0137] S1102, setting the parameters and variable range of the NSGA-III genetic algorithm;

[0138] S1104, randomly generate the initial population;

[0139] S1106, selecting an operation based on a reference point;

[0140] S1108, performing crossover and mutation operations to obtain a progeny population;

[0141] S1110, merge the parent population and the offspring population;

[0142] S1112, performing non-dominated sorting according to the objective function value;

[0143] S1114, select the better individuals to enter the next generation population;

[0144] S1116, whether the number of iterations has been reached; specifically, if not, execute S1118, if yes, the process ends;

[0145] S1118, increase the number of iterations by 1, and return to S1104.

[0146] The power device package structure optimization design method based on response surface and NSGA-Ⅲ algorithm proposed in this embodiment has good optimization effect. This method combines finite element simulation method and mathematical analysis method. Compared with the simple simulation optimization method, the efficiency is greatly improved, avoiding blind large-scale simulation and shortening the design cycle; compared with the simple mathematical analysis method, there is no need to derive the optimization target expression according to the subject theory and model structure, which greatly simplifies the optimization process, reduces the professional threshold of users, is applicable to different models, has good versatility, and can better reflect the details of the power device structure optimization design. In addition, although two optimization targets are selected in this embodiment, the power device package structure optimization design method based on response surface and NSGA-Ⅲ algorithm proposed in the present invention has better optimization effect on high-dimensional target optimization (i.e., three or more optimization targets). This method can solve the problem that the previous method has unsatisfactory optimization effect on high-dimensional target optimization problems, and is applicable to many multi-objective optimization problems with 2 to 15 targets, and can achieve satisfactory results.

[0147] In this specification, the term "plurality" means two or more than two, unless otherwise clearly defined. The terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense. For example, "connected" can be a fixed connection, a detachable connection, or an integral connection; "connected" can be a direct connection or an indirect connection through an intermediate medium. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances.

[0148] In the description of this specification, the description of the terms "one embodiment", "some embodiments", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0149] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A power device structure design method based on response surface and NSGA-Ⅲ algorithm, characterized in that: include: Obtaining packaging form, packaging structure and chip layout information of the power device; Determining the size information of the packaging structure and the size information of the chip layout; Get the design goals; Determine design variables and fixed parameters in the size information of the package structure and the size information of the chip layout according to the design goal; According to the design variables and the fixed parameters, a design mathematical model of the design target is obtained by fitting using the response surface method; According to the design mathematical model, the design mathematical model is calculated using the NSGA-III algorithm to obtain a solution set of the design mathematical model, so as to design a power device structure according to the solution set; Before the step of determining the design variables and fixed parameters in the size information of the package structure and the size information of the chip layout according to the design target, the method further includes: Determining variable size parameters in the size information of the package structure and the size information of the chip layout; Determining a numerical variation range of the variable size parameter; The step of determining the design variables and fixed parameters in the size information of the package structure and the size information of the chip layout according to the design goal specifically includes: After determining the value variation range of the variable size parameter, obtaining the size parameter as the design variable and the size parameter as the fixed parameter according to the influence degree of the variable size parameter on the design target; The influence degree of the variable size parameter on the design target is obtained through single variable sensitivity analysis and / or manually based on experience; When the parameter values ​​of the design variable and the fixed parameter change in the same manner, the influence of the design variable on the design target is greater than the influence of the fixed parameter on the design target.

2. The power device structure design method based on response surface and NSGA-III algorithm according to claim 1, characterized in that: The step of obtaining the packaging form, packaging structure and chip layout information of the power device specifically includes: Obtain the packaging form of the actual power device to be structurally designed; Acquire the packaging structure of the actual power device and the dimensions of each layer of the packaging structure; The number of chips, chip positions and distances between adjacent chips of the actual power device are obtained.

3. The power device structure design method based on response surface and NSGA-III algorithm according to claim 1 is characterized in that: The design target includes at least one or more of chip maximum stress, chip maximum temperature, and module inductance.

4. The power device structure design method based on response surface and NSGA-III algorithm according to claim 1, characterized in that: The step of fitting the design mathematical model of the design target by using the response surface method according to the design variables and the fixed parameters specifically includes: Determine the variable experimental combination by using the response surface method according to the design goal, the design variables, and the numerical variation range of the variable size parameter; Perform finite element simulation according to the variable experimental combination to obtain simulation results; The regression coefficient is calculated by using the response surface method according to the simulation results; Obtaining the objective function and constraint function of the design objective according to the regression coefficient; The design mathematical model is obtained according to the objective function and the constraint function.

5. A power device structure design system based on response surface and NSGA-Ⅲ algorithm, characterized in that: The power device structure design method based on the response surface and NSGA-III algorithm according to any one of claims 1 to 4 is adopted, and the power device structure design system based on the response surface and NSGA-III algorithm comprises: An information acquisition module, used to acquire the packaging form, packaging structure and chip layout information of the power device, determine the size information of the packaging structure and the size information of the chip layout, acquire the design goal, and determine the design variables and fixed parameters in the size information of the packaging structure and the size information of the chip layout according to the design goal; A calculation module is used to obtain a design mathematical model of the design target by fitting the design variables and the fixed parameters using the response surface method, and to calculate the design mathematical model using the NSGA-III algorithm based on the design mathematical model to obtain a solution set of the design mathematical model for designing a power device structure according to the solution set.

6. A power device structure design system based on response surface and NSGA-Ⅲ algorithm, characterized in that: include: A memory, a processor, and a program stored in the memory and executable on the processor, wherein when the program is executed by the processor, the steps defined in the power device structure design method based on the response surface and NSGA-III algorithm as described in any one of claims 1 to 4 are implemented.

7. A readable storage medium, characterized in that: Programs and / or instructions are stored thereon, and when the programs and / or instructions are executed by a processor, the steps of the power device structure design method based on the response surface and NSGA-III algorithm as described in any one of claims 1 to 4 are implemented.