Electromagnetic field analysis simulation system and method based on chip packaging

Abstract

The invention relates to the technical field of chip packaging electromagnetic field analysis, in particular to a simulation system and method based on chip packaging electromagnetic field analysis, and the method comprises the steps: 1, obtaining chip packaging parameters from a database, and constructing an electromagnetic field model of a chip; the chip packaging parameters comprise packaging structure parameters and packaging material parameters; 2, analyzing the electromagnetic field model of the chip, and obtaining distribution characteristics of an electric field and a magnetic field in the chip and current distribution characteristics through an analysis result; 3, learning the relation between the magnetic field distribution characteristics and the current distribution characteristics through the prediction model, and predicting the current distribution under different magnetic field distribution conditions; step 4, adjusting a chip packaging parameter based on a prediction result, optimizing the packaging parameter through a genetic algorithm, and searching an optimal packaging parameter; and step 5, carrying out simulation verification on the optimized chip packaging parameters.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic field analysis technology for chip packaging, and particularly to a simulation system and method for electromagnetic field analysis based on chip packaging. Background Technology

[0002] Electromagnetic field analysis is crucial during chip packaging. At high frequencies or under AC conditions, the wavelength of the signal is on the same order of magnitude as the size of the package leads, and signal pulses exhibit significant fluctuations on the interconnects. Therefore, the analysis of the electromagnetic characteristics of interconnects, package structures, and especially interconnect systems plays an increasingly important role in the analysis and design of high-speed integrated circuit systems. Electromagnetic field analysis in chip packaging helps optimize circuit design and reduce the generation and propagation of electromagnetic interference (EMI). For example, through proper pin assignment and decoupling capacitor design, EMI can be effectively controlled and signal integrity improved.

[0003] The distribution of electromagnetic fields affects the distribution of current. If the heat generated by the current accumulates in a certain area, the temperature in that area will rise, thus affecting the performance and reliability of the device. To address this, we propose a method to predict and optimize the current distribution by analyzing the electromagnetic field of the chip package, thereby improving the chip's thermal management. Summary of the Invention

[0004] This invention is based on the analysis of the electromagnetic field of chip packaging, combined with chip packaging parameters, to establish the relationship between chip packaging parameters and chip internal current density. By optimizing the current distribution, overheating inside the chip is avoided, and the optimization results are verified by simulation.

[0005] The technical solution proposed in this invention is: a chip packaging electromagnetic field analysis and simulation method, the method comprising:

[0006] Step 1: Obtain chip packaging parameters from the database and construct the electromagnetic field model of the chip; the chip packaging parameters include packaging structure parameters and packaging material parameters;

[0007] Step 2: Analyze the electromagnetic field model of the chip, and obtain the distribution characteristics of the electric and magnetic fields and the current distribution characteristics inside the chip through the analysis results;

[0008] Step 3: Learn the relationship between magnetic field distribution characteristics and current distribution characteristics through the prediction model, and predict the current distribution under different magnetic field distribution conditions;

[0009] Step 4: Adjust the chip packaging parameters based on the prediction results, and optimize the packaging parameters using a genetic algorithm to find the optimal packaging parameters;

[0010] Step 5: Perform simulation verification on the optimized chip packaging parameters.

[0011] Preferably, the packaging material parameters include, but are not limited to, the dielectric constant, permeability, and resistivity of the packaging material; the packaging structure parameters include, but are not limited to, the width, length, and thickness of the chip package.

[0012] Preferably, the step of obtaining chip packaging parameters from the database and constructing the chip's electromagnetic field model includes the following steps:

[0013] The chip packaging structure parameters are obtained from the database. Using the chip packaging structure parameters, a chip packaging structure model is established using the finite element method (FEM). The chip packaging structure model is then meshed to obtain multiple finite element mesh elements.

[0014] Retrieve chip packaging material parameters from the database, and define material properties based on the packaging material parameters, including setting air node constant, permeability and dielectric constant;

[0015] Define boundary conditions, including Dirichlet boundary conditions or Neumann boundary conditions;

[0016] For each finite element mesh, a first-order polynomial basis function is selected as the interpolation function, and the stiffness matrix of each finite element mesh is calculated.

[0017] Assemble all stiffness matrices into a global stiffness matrix, and apply boundary conditions to modify the global stiffness matrix.

[0018] Preferably, the analysis of the electromagnetic field model of the chip, and the acquisition of the distribution characteristics of the electric and magnetic fields and the current distribution characteristics inside the chip through the analysis results, includes the following steps:

[0019] Importing Maxwell's equations:

[0020] ;

[0021] in, and These represent electric field strength, magnetic induction intensity, magnetic field strength, current density, electric displacement vector, and charge density, respectively.

[0022] Transform the partial differential equations in Maxwell's equations into weaker form equations:

[0023] ;in, These represent the test functions, These represent scalar test functions;

[0024] Obtain magnetic permeability, electrical conductivity, and current density;

[0025] Apply boundary conditions;

[0026] The implicit Euler method is used for time-stepping loops;

[0027] The magnetic field strength is obtained by solving the weak form equation, and the magnetic field streamline diagram is plotted using matplotlib.

[0028] Calculate the current density distribution using Ampere's circuital law;

[0029] Plot current density contour lines using matplotlib.

[0030] Preferably, the step of learning the relationship between magnetic field distribution characteristics and current distribution characteristics through a prediction model to predict the current distribution under different magnetic field distribution conditions includes the following steps:

[0031] A prediction model is constructed to predict the characteristics of electric field distribution; the prediction model is a multilayer perceptron neural network, including an input layer, a hidden layer and an output layer;

[0032] ReLU is used as the activation function for the hidden layer, a linear activation function is used for the output layer, and mean squared error is used as the loss function.

[0033] Historical magnetic field streamline diagrams are obtained from the database. Key magnetic field features are extracted from the historical magnetic field streamline diagrams and standardized to form a set of key magnetic field features, including magnetic field strength and magnetic field direction.

[0034] Historical current density contour maps are obtained from the database. Key electric field features are extracted from the historical current density contour maps and standardized to form a set of key electric field features, which includes current density and current direction.

[0035] The key feature sets of magnetic fields and electric fields are combined to form the original dataset 1. The original dataset 1 is then divided into training set 1 and test set 1.

[0036] Train the prediction model using the training set, and test the prediction model using the test set.

[0037] The trained prediction model is used to predict the electric field characteristics of the newly input magnetic field features, and the actual electric field feature values ​​are obtained by inverse standardization.

[0038] Preferably, it also includes establishing a relational model to represent the relationship between encapsulation parameters and electric field characteristics, wherein the relational model is a random forest regression model;

[0039] Obtain the chip's packaging parameters from the database. These packaging parameters include, but are not limited to, the dielectric constant and resistivity of the chip packaging material, as well as the chip packaging size.

[0040] Obtain electric field characteristic data and corresponding packaging parameters of chips with different packages from the database;

[0041] Key packaging features are extracted from the packaging parameters, and after standardization of the key packaging features, a packaging parameter feature set is formed.

[0042] The encapsulated feature set and the key electric field feature set constitute the second original dataset.

[0043] The original dataset is divided into two sets: a training set and a test set.

[0044] The relation model is trained using training set 2, and the performance of the relation model is tested using test set 2.

[0045] Set a mean squared error threshold and a coefficient of determination threshold, and evaluate the performance of the relational model by whether the requirements of the corresponding thresholds are met.

[0046] The electric field characteristics are the target variable, the encapsulation parameters are the input variables, and the importance of the encapsulation parameters is output using a trained relational model.

[0047] Create a bar chart to show the order of importance.

[0048] Preferably, the step of adjusting the chip packaging parameters based on the prediction results and optimizing the packaging parameters using a genetic algorithm to find the optimal packaging parameters includes the following steps:

[0049] Obtain the feature values ​​of the two most important encapsulation parameters;

[0050] Establish applicability function This is used to represent the relationship between the current density at the location to be improved within the chip and the dielectric constant of the packaging material and the chip package thickness. Indicates the current density value. These represent the dielectric constant of the packaging material and the chip package thickness, respectively.

[0051] Using a genetic algorithm to find... The minimum optimal packaging parameters are used to optimize current distribution and improve chip thermal management.

[0052] The obtained optimal packaging parameters and corresponding current density values ​​are transferred to a database for storage.

[0053] Preferably, the step of re-simulating and verifying the packaging parameters of the optimized chip includes the following steps:

[0054] The optimal chip packaging parameters and corresponding current density values ​​are obtained from the database to construct the electromagnetic field model of the chip;

[0055] The electromagnetic field model of the chip is analyzed to obtain the contour map of the current density inside the chip;

[0056] Extract the current density values ​​of the locations to be improved from the newly acquired current density contour map. ,if, If the optimization is successful, it is considered effective; otherwise, it is considered ineffective, and step 4 is repeated.

[0057] The present invention also provides a chip-package-based electromagnetic field analysis and simulation system, which is used to execute the chip-package-based electromagnetic field analysis and simulation method.

[0058] The present invention also provides a computer-readable storage medium storing a computer program, which is executed by a processor to implement the chip-package-based electromagnetic field analysis and simulation method.

[0059] The beneficial effects of this invention are:

[0060] 1. This invention first extracts the structural and material parameters of the chip package from a database. Using these parameters, a three-dimensional electromagnetic field model is constructed to simulate the electromagnetic environment of the chip within the package. Then, the constructed electromagnetic field model is analyzed using the finite element method (FEM). By solving Maxwell's equations, the distribution of the electric and magnetic fields and the current density inside the chip are obtained.

[0061] 2. This invention uses machine learning techniques (such as neural networks) to learn the relationship between magnetic field distribution characteristics and current distribution. A predictive model is established to predict the corresponding current distribution based on different magnetic field configurations. A genetic algorithm (GA) is used in conjunction with the predictive model to optimize chip packaging parameters. An optimal set of packaging parameters is found so that, under given magnetic field conditions, the current distribution inside the chip achieves the expected design goal. This reduces the current density at critical locations inside the chip, preventing overheating and achieving an auxiliary thermal management effect. Attached Figure Description

[0062] Figure 1 This is a flowchart of the electromagnetic field analysis and simulation method based on chip packaging according to the present invention. Detailed Implementation

[0063] The following description is intended to disclose the present invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art. The basic principles of the invention defined in the following description can be applied to other embodiments, modifications, improvements, equivalents, and other technical solutions that do not depart from the spirit and scope of the invention.

[0064] It is understood that the term "a" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0065] refer to Figure 1 The technical solution provided by this invention is: a chip packaging electromagnetic field analysis and simulation method, comprising the following steps:

[0066] Step 1: Obtain chip packaging parameters from the database and construct the electromagnetic field model of the chip; the chip packaging parameters include packaging structure parameters and packaging material parameters; the packaging material parameters include, but are not limited to, the dielectric constant, permeability and resistivity of the packaging material; the packaging structure parameters include, but are not limited to, the width, length and thickness of the chip package.

[0067] The electromagnetic field model of the chip includes the following steps:

[0068] The chip packaging structure parameters are obtained from the database. Using the chip packaging structure parameters, a chip packaging structure model is established using the finite element method (FEM). The chip packaging structure model is then meshed to obtain multiple finite element mesh elements. In this embodiment, the chip packaging structure model is divided into 32*32 finite element elements using the mesh.

[0069] Retrieve chip packaging material parameters from the database, and define material properties based on these parameters, including setting air node constant, permeability, and dielectric constant.

[0070] Define boundary conditions, including Dirichlet boundary conditions or Neumann boundary conditions; in this embodiment, Dirichlet boundary conditions are set, that is, the potential of the boundary is 0.

[0071] For each finite element mesh, a first-order polynomial basis function is selected as the interpolation function to calculate the stiffness matrix of each mesh. Specifically: the first-order polynomial is defined as follows: ,in The coefficients are related to the nodes, where each node is a vertex of a finite element.

[0072] The formula for calculating the local stiffness matrix is: ,in, Indicates the first A finite element mesh;

[0073] Assemble all stiffness matrices into a global stiffness matrix, and apply boundary conditions to modify the global stiffness matrix.

[0074] Step 2: Analyze the electromagnetic field model of the chip to obtain the distribution characteristics of the electric and magnetic fields and the current distribution characteristics inside the chip; specifically, this includes the following steps:

[0075] Includes the following steps:

[0076] Importing Maxwell's equations:

[0077] ;

[0078] in, and These represent electric field strength, magnetic induction intensity, magnetic field strength, current density, electric displacement vector, and charge density, respectively.

[0079] Transform the partial differential equations in Maxwell's equations into weaker form equations:

[0080] ;in, These represent the test functions, These represent scalar test functions;

[0081] Obtain the permeability, conductivity, and current density; apply boundary conditions; use the implicit Euler method for time-stepping loops, setting the step size to 0.01s, the time to T, and the number of loops to T / 0.01.

[0082] By solving the weak form equations, the magnetic field strength and electric field strength are obtained, and the streamplot function of the matplotlib library is used to draw the magnetic field streamline diagram.

[0083] The current density distribution is calculated using Ampere's circuital law; the contour plot of the current density is drawn using the contourf function of the matplotlib library.

[0084] Step 3: Learn the relationship between magnetic field distribution characteristics and current distribution characteristics through a prediction model to predict the current distribution under different magnetic field distribution conditions; specifically, this includes the following steps:

[0085] A prediction model is constructed to predict the characteristics of electric field distribution; the prediction model is a multilayer perceptron neural network (MLP), including an input layer, a hidden layer and an output layer;

[0086] ReLU is used as the activation function for the hidden layer, a linear activation function is used for the output layer, and mean squared error is used as the loss function.

[0087] Historical magnetic field streamline maps are obtained from the database, and key magnetic field features, including magnetic field strength and direction, are extracted from them. After standardization, a set of key magnetic field features is constructed.

[0088] Historical current density contour maps are obtained from the database. Key electric field features are extracted from the historical current density contour maps and standardized to form a set of key electric field features, which includes current density and current direction.

[0089] The key feature sets of magnetic fields and electric fields are combined to form the original dataset 1. The original dataset 1 is then divided into training set 1 and test set 1.

[0090] Train the prediction model using the training set, and test the prediction model using the test set.

[0091] The trained prediction model is used to predict the electric field characteristics of the newly input magnetic field features, and the actual electric field feature values ​​are obtained by inverse standardization.

[0092] Before improving current density distribution by modifying packaging parameters, it is necessary to determine the relationship between packaging parameters and electric field characteristics. The specific process is as follows:

[0093] A relational model is established based on a random forest regression model; the chip packaging parameters are obtained from the database, including but not limited to the dielectric constant, resistivity, and chip packaging size of the chip packaging material; the electric field characteristic data and corresponding packaging parameters of different packaged chips are obtained from the database; key packaging features are extracted from the packaging parameters, and after standardization of the key packaging features, a packaging parameter feature set is formed.

[0094] The encapsulated feature set and the key electric field feature set constitute the second original dataset. The second original dataset is then divided into a training set and a test set.

[0095] The relation model is trained using training set 2, and its performance is tested using test set 2.

[0096] Set a mean squared error threshold and a coefficient of determination threshold, and evaluate the performance of the relational model by whether the requirements of the corresponding thresholds are met.

[0097] Electric field characteristics are the target variable, and encapsulation parameters are the input variables. A trained relational model is used to output an importance score for the encapsulation parameters, which are then ranked according to their importance scores. A bar chart of the importance scores is plotted to show the order of importance.

[0098] Step 4: Adjust the chip packaging parameters based on the prediction results, and optimize the packaging parameters using a genetic algorithm to find the optimal packaging parameters; specifically, this includes the following steps:

[0099] Obtain the feature values ​​of the two most important encapsulation parameters;

[0100] Establish applicability function This is used to represent the relationship between the current density at the location to be improved within the chip and the dielectric constant of the packaging material and the chip package thickness. Indicates the current density value. These represent the dielectric constant of the packaging material and the chip package thickness, respectively.

[0101] Using a genetic algorithm to find... The minimum optimal packaging parameters are used to optimize current distribution and improve chip thermal management.

[0102] The obtained optimal packaging parameters and corresponding current density values ​​are transferred to a database for storage.

[0103] Step 5: Perform simulation verification on the optimized chip packaging parameters, which includes the following steps:

[0104] The optimal chip packaging parameters and corresponding current density values ​​are obtained from the database to construct the electromagnetic field model of the chip;

[0105] The simulation program is run to analyze the electromagnetic field model of the chip and obtain contour maps of the current density inside the chip. Analysis of these contour maps identifies areas with excessively high current density, which are the locations requiring optimization. A unique identification tag is added to each identified area with high current density, and this tag is sent to a host computer. Technicians use the corresponding identification tag to determine whether the area needs improvement, thus identifying the areas requiring improvement. Specifically, preset current density thresholds can be used to determine which areas have excessively high current density.

[0106] Extract the current density values ​​of the locations to be improved from the newly acquired current density contour map. This serves as the basis for evaluating the effectiveness of the optimization.

[0107] if, If the optimization is successful, it is considered effective; otherwise, it is considered ineffective, and step 4 above is repeated.

[0108] After optimizing the packaging parameters, simulations were performed again to verify whether the optimized parameters reduced the current density at the corresponding location (the location to be improved). The difference between the two values ​​was within the set threshold range, indicating that the optimized packaging parameters were effective in reducing the current density. By optimizing the current distribution, overheating inside the chip can be effectively avoided, thereby protecting the safe operation of critical chips such as the central processing unit (CPU) and graphics processing unit (GPU).

[0109] The simulation tools used in this embodiment are SPICE and MATLAB Simulink.

[0110] The present invention also provides a chip-package-based electromagnetic field analysis and simulation system, which is used to execute the chip-package-based electromagnetic field analysis and simulation method.

[0111] The present invention also provides a computer-readable storage medium storing a computer program, which is executed by a processor to implement the chip-package-based electromagnetic field analysis and simulation method.

[0112] The processes described above with reference to the flowcharts in the embodiments disclosed in this invention can be implemented as computer software programs. The embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), it performs the functions defined in the methods of this application. It should be noted that the computer-readable medium described above in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wire segments, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless segments, wire segments, optical fibers, RF, etc., or any suitable combination thereof.

[0113] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0114] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are merely examples and do not limit the present invention. The purpose of the present invention has been fully and effectively achieved. The functions and structural principles of the present invention have been shown and explained in the embodiments. Without departing from the stated principles, the implementation of the present invention may have any variations or modifications.

Claims

1. Based on the chip packaging electromagnetic field analysis simulation method, it is characterized by: The method comprises: Step 1, obtaining chip packaging parameters from a database and constructing an electromagnetic field model of the chip; the chip packaging parameters include packaging structure parameters and packaging material parameters; Step 2: Analyze the electromagnetic field model of the chip, and obtain the distribution characteristics of the electric field and magnetic field and the current distribution characteristics inside the chip through the analysis results; Step 3: Learn the relationship between magnetic field distribution characteristics and current distribution characteristics through the prediction model, and predict the current distribution under different magnetic field distribution conditions; Step 4: Adjust chip packaging parameters based on the prediction results, and optimize the packaging parameters through genetic algorithms to find the optimal packaging parameters; Step 5: Simulate and verify the optimized chip packaging parameters.

2. The chip packaging electromagnetic field analysis and simulation method according to claim 1 is characterized in that: The packaging material parameters include but are not limited to the dielectric constant, magnetic permeability and resistivity of the packaging material; the packaging structure parameters include but are not limited to the width, length and thickness of the chip package.

3. The chip packaging electromagnetic field analysis and simulation method according to claim 2 is characterized in that: The step of obtaining chip packaging parameters from a database and constructing an electromagnetic field model of the chip includes the following steps: Obtaining chip packaging structure parameters from a database, using the chip packaging structure parameters, establishing a chip packaging structure model through a finite element method (FEM), and meshing the chip packaging structure model to obtain a plurality of finite element mesh units; Obtain chip packaging material parameters from the database, and define material properties based on the packaging material parameters, including setting air node constants, magnetic permeability, and dielectric constant; Define boundary conditions, including Dirichlet boundary conditions or Neumann boundary conditions; A first-order polynomial basis function is selected as an interpolation function for each finite element mesh, and the stiffness matrix of each finite element mesh is calculated; All stiffness matrices are assembled into a global stiffness matrix, and boundary conditions are applied to modify the global stiffness matrix.

4. The chip packaging electromagnetic field analysis and simulation method according to claim 3 is characterized in that: The electromagnetic field model of the chip is analyzed, and the distribution characteristics of the electric field and magnetic field and the current distribution characteristics inside the chip are obtained through the analysis results, including the following steps: Import Maxwell's equations: ； in, and They represent electric field intensity, magnetic induction intensity, magnetic field intensity, current density, electric displacement vector and charge density respectively; Convert the partial differential equations in Maxwell's equations to weak form: ;in, They represent the test functions, They represent scalar test functions respectively; Obtain magnetic permeability, electrical conductivity, and current density; Apply boundary conditions; The implicit Euler method is used for time stepping loop; By solving the weak form equation, the magnetic field strength is obtained, and the magnetic field streamline diagram is drawn using matplotlib; Calculate the current density distribution by Ampere's circuit theorem; The current density contour map is drawn using matplotlib.

5. The chip packaging electromagnetic field analysis and simulation method according to claim 4 is characterized in that: The method of learning the relationship between the magnetic field distribution characteristics and the current distribution characteristics through the prediction model and predicting the current distribution under different magnetic field distribution conditions includes the following steps: Constructing a prediction model for predicting electric field distribution characteristics; the prediction model is a multi-layer perceptron neural network, including an input layer, a hidden layer and an output layer; Use ReLU as the activation function of the hidden layer, linear activation function of the output layer, and mean square error as the loss function; Obtain a historical magnetic field streamline map from a database, extract key magnetic field features from the historical magnetic field streamline map, and form a magnetic field key feature set after standardization, wherein the magnetic field key features include magnetic field intensity and magnetic field direction; Obtain a historical current density contour map from a database, extract key electric field features from the historical current density contour map, and form a set of key electric field features after standardization, wherein the key electric field features include current density and current direction; The magnetic field key feature set and the electric field key feature set are combined to form an original data set 1, and the original data set 1 is divided into a training set 1 and a test set 1; Use the training set to train the prediction model, and use the test set to test the performance of the prediction model; The trained prediction model is used to predict the electric field characteristics of the newly input magnetic field characteristics, and the actual electric field characteristic values ​​are obtained by denormalization.

6. The chip packaging electromagnetic field analysis and simulation method according to claim 1, characterized in that: It also includes establishing a relationship model for representing the relationship between packaging parameters and electric field characteristics, wherein the relationship model is a random forest regression model; Acquire chip packaging parameters from a database, wherein the packaging parameters include but are not limited to dielectric constant and resistivity of chip packaging materials and chip packaging size; Obtain electric field characteristic data and corresponding packaging parameters of different packaged chips from the database; Extract key packaging features from packaging parameters, and form a packaging parameter feature set after standardizing the key packaging features; The package feature set and the electric field key feature set constitute the original data set 2; Divide the original data set into two parts: training set 2 and test set 2; The relational model is trained using training set 2, and the performance of the relational model is tested using test set 2; Set the mean square error threshold and the determination coefficient threshold, and evaluate the performance of the relationship model by whether the corresponding threshold requirements are met; The electric field characteristics are used as the target variables, the packaging parameters are used as the input variables, and the importance of the packaging parameters is output using the trained relationship model; Draw a bar graph to show the order of importance.

7. The chip packaging electromagnetic field analysis and simulation method according to claim 6 is characterized in that: The method of adjusting chip packaging parameters based on the prediction results and optimizing the packaging parameters by genetic algorithm to find the optimal packaging parameters includes the following steps: Get the package parameter feature values ​​ranked in the top two in terms of importance; Building a fitness function , which is used to express the relationship between the current density at the location to be improved inside the chip and the dielectric constant of the packaging material and the chip packaging thickness, where: represents the current density value, They represent the dielectric constant of the packaging material and the chip packaging thickness respectively; Genetic algorithm is used to find Minimum optimal packaging parameters to optimize current distribution and improve chip thermal management; The obtained optimal packaging parameters and corresponding current density values ​​are transmitted and stored in a value database.

8. The chip packaging electromagnetic field analysis and simulation method according to claim 7 is characterized in that: The re-simulation verification of the packaging parameters of the optimized chip includes the following steps: Obtain the optimal chip packaging parameters and corresponding current density values ​​from the database and construct the electromagnetic field model of the chip; Analyze the electromagnetic field model of the chip and obtain the contour map of the current density inside the chip; Extract the current density value of the location to be improved from the newly obtained current density contour map ,if, If yes, the optimization is considered effective; otherwise, the optimization is considered invalid and step 4 is repeated.

9. Based on the chip packaging electromagnetic field analysis simulation system, it is characterized by: The system is used to execute the chip packaging electromagnetic field analysis simulation method described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is executed by a processor to implement the chip packaging electromagnetic field analysis simulation method according to any one of claims 1 to 8.

Images