A time-domain electromagnetic simulation method based on distributed summation method and simultaneous approximation term

By using the distribution summing method and simultaneous approximation term methods in time domain electromagnetic simulation, the step error and numerical instability of sub-grid technology in fine structure processing are solved, and efficient and accurate Electromagnetic field simulation ensures long-term stability of system energy.

CN118551598BActive Publication Date: 2024-11-22BEIHANG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410242583.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-03-04
Publication Date
2024-11-22
Estimated Expiration
2044-03-04

AI Technical Summary

Technical Problem

The time-domain finite difference method has step errors when dealing with fine structures and surface structures, resulting in low modeling accuracy and long simulation time. Traditional sub-mesh technology is prone to numerical instability in long-term simulations.

Method used

The time-domain electromagnetic simulation method based on the distribution summation method and simultaneous approximation terms are adopted. By using uniform meshes in different regions, the algorithm is designed to maintain stability, and the electromagnetic field values ​​are calculated separately in the main mesh and sub-grid areas, and projection is introduced. Operators regulate energy changes and ensure long-term stability of system energy.

Benefits of technology

It improves the calculation efficiency and accuracy, ensures the system energy stability when dealing with the electromagnetic field simulation problem of fine structures, and improves the accuracy and reliability of simulation results.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118551598B_ABST
    Figure CN118551598B_ABST
Patent Text Reader

Abstract

The present invention discloses a time-domain electromagnetic simulation method based on the distribution summation method and the simultaneous approximation term, comprising the following steps: S1. defining the simulation target area, discretizing the main grid, and giving the known information of each grid point; S2. local encryption discretization of the sub-grid: on the basis of the main grid, further dividing each main grid point to form smaller sub-grid points; S3. calculating the electromagnetic field at each grid; S4. calculating the electromagnetic field under boundary conditions; S5. selecting the time step and the number of iterations; S6. in each iteration process, repeating steps S3-S4, obtaining the electromagnetic field calculation results in each iteration process, and exporting them as the final time-domain electromagnetic simulation results. This application is extended by the regional uniform grid, and the design algorithm can maintain stability in different regions and different grid sizes, so as to meet the needs of practical problems.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of electromagnetic simulation, and in particular to a time domain electromagnetic simulation method based on a distributed summation method and a simultaneous approximation term. Background Art

[0002] As an important branch of computational electromagnetics, the finite-difference time-domain method is simple, efficient and widely used. However, since the finite-difference time-domain method usually uses a uniform hexahedral grid with a fixed size, it has step errors and cannot accurately and efficiently process fine structures and curved structures:

[0003] (1) Due to the use of regular hexahedral meshes, there are step errors. Therefore, in order to improve modeling accuracy, smaller meshes are often used, but this will increase memory consumption and increase simulation time;

[0004] (2) Subgrid technology uses fine grids locally, which can appropriately improve the solution efficiency based on the poor modeling accuracy of complex structures and the existence of large step approximation errors. However, traditional subgrid technology is prone to numerical instability during long-term simulations. Summary of the invention

[0005] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a time-domain electromagnetic simulation method based on the distributed summation method and the simultaneous approximation term, which is expanded by a regional uniform grid. The design algorithm can maintain stability in different regions and different grid sizes to meet the needs of practical problems.

[0006] The object of the present invention is achieved through the following technical solution: A time domain electromagnetic simulation method based on distributed summation method and simultaneous approximation term, comprising the following steps:

[0007] S1. Define the simulation target area, discretize the main grid, and give the known information of each grid point;

[0008] The step S1 comprises:

[0009] S101. Determine the simulation area: Given a target object to be simulated, and set the size range of the entire rectangular simulation area containing the target object: length is x, width is y;

[0010] S102. Determine the boundary condition type:

[0011] The boundary condition type is a closed boundary condition, which means that when electromagnetic waves in the simulation area propagate to the boundary of the simulation area, reflection will be directly generated;

[0012] S103. Main grid discretization: According to the wavelength λ of the simulated electromagnetic wave, the resolution of spatial discretization is determined, and a single main grid size of (λ / 20)×(λ / 20) square grid is set; after the grid covers the entire calculation area, discrete grid points are set in the simulation area. These grid points will be used to record the values ​​of the electric and magnetic fields. The number of grid points in the x direction is x / (λ / 20)+1, and the number of grid points in the y direction is y / (λ / 20)+1. For each grid point, the following information is included: grid point position (x ij ,y ik ), the electric field strength E at the grid point i , magnetic field strength H i ;

[0013] S104 determines the material distribution: determines the materials of each part in the simulation area, describes the material parameters corresponding to each grid point, and assigns values. The assigned parameters include: the material's magnetic permeability μ, the material's dielectric constant and conductivity ε, and the material's electrical conductivity σ ;

[0014] S105. Setting initial conditions: At the beginning of the simulation, it is necessary to specify the initial conditions of the electric field and the magnetic field, including the distribution of the electric field and the magnetic field, that is, specifying the magnetic field and the electric field at each grid point;

[0015] Specify the information of the excitation source. If point current source excitation is used, the following information is included: excitation source position (x i ,y i ), excitation amplitude U, excitation source waveform change f, excitation source wavelength λ.

[0016] S2. Sub-grid local refinement discretization: Based on the main grid, each main grid point is further divided into smaller sub-grid points;

[0017] In step S2, the fine structure position where the resolution needs to be increased is set as region (x s1 ,y s1 )×(x s2 ,y s2 ), the grid refinement ratio is 2 times, then the number of grid points in the sub-grid x direction is: (x s2 -x s1 ) / (λ / 20 / 2)+1, the number of grid points in the y direction is: (y s2 -y s1 ) / (λ / 20 / 2)+1, the information contained in each grid point is consistent with the main grid point.

[0018] S3. Based on the distributed summation operator that maintains the long-term stability of energy, the Maxwell equations corresponding to the space are defined and set so that the corresponding semi-discrete equations can describe the changes of the electromagnetic field in the space, and then the obtained Maxwell equations are used to calculate the electromagnetic field at each grid;

[0019] The step S3 comprises:

[0020] S301. Given a distributed summation operator that maintains long-term energy stability, including: matrix a y , and |a x 、a| x and

[0021] in,

[0022]

[0023]

[0024]

[0025]

[0026] S302. Using these one-dimensional distribution sum discretization matrices, construct a two-dimensional distribution sum discretization matrix as follows:

[0027]

[0028]

[0029] I x is the unit diagonal matrix corresponding to the number of grid points in the x direction of the target area, I y The unit diagonal matrix corresponding to the number of grid points in the y direction of the target area;

[0030] S303. The two-dimensional Maxwell equations are in the following form:

[0031]

[0032] Among them, H x , H y are the magnetic field components in the x and y directions respectively, E z is the magnetic field component in the z direction, μ is the magnetic permeability of the mesh material, and ε is the dielectric constant of the mesh material;

[0033] Based on the two-dimensional distribution sum discretization matrix, the Maxwell equations corresponding to the space are defined and set, and the two-dimensional Maxwell equations are discretized into:

[0034]

[0035] S304. When the target area is the main grid and sub-grid area, according to steps S302 to S303, the corresponding two-dimensional distribution sum discretization matrix is ​​calculated respectively, and then the two-dimensional Maxwell equations of the main grid and sub-grid area are obtained respectively, so as to calculate the electromagnetic field value of each grid point of the main grid and sub-grid area; the electromagnetic field value mainly calculated for each grid point includes H x , H y and E z .

[0036] S4. Setting of boundary conditions based on simultaneous approximation terms: Based on the boundary conditions of the simultaneous approximation terms with adjusted energy, each boundary condition is set and the energy is regulated so as to achieve long-term stability, and the electromagnetic field under the boundary conditions is calculated;

[0037] The step S4 comprises:

[0038] S401. On the premise of discretization of the two-dimensional Maxwell equations in step S3, adding boundary conditions of simultaneous approximation terms to perform energy regulation;

[0039] For closed boundary conditions, after adding the boundary conditions of the simultaneous approximation term, the obtained two-dimensional Maxwell equations are:

[0040]

[0041] Among them, σ L , σ R , σ B and σ T is a scalar;

[0042] Take σ L =-1,σ R =1,σ B =-1,σ T =1, so that under the closed boundary condition, the rate of change of the electromagnetic field energy in the region with time is always 0, thereby achieving long-term stability of energy by simultaneously approximating the boundary conditions;

[0043] S402. After adding the boundary conditions of the simultaneous approximation term for the main grid and sub-grid boundaries, the obtained two-dimensional Maxwell equations are:

[0044] The main grid boundaries are:

[0045]

[0046] The subgrid boundaries are:

[0047]

[0048] in, is an unknown scalar parameter;

[0049] Pick All are -1 / 2, so that the electromagnetic energy of the main grid and the sub-grid interacts, and the time conversion rate is always 0, maintaining the long-term stability of energy;

[0050] Among them, the interpolation matrix and They are:

[0051]

[0052]

[0053] S403. According to steps S401 to S402, the electromagnetic fields of the closed boundary, the main grid boundary and the sub-grid boundary are calculated.

[0054] S5. Select the time step and number of iterations: Set the time step length and the total simulation time. The single time interval t is determined by the subgrid size and is The number of iterations is the ratio of the total computation time T to a single time interval t;

[0055] S6. In each iteration process, steps S3 to S4 are repeatedly executed to obtain the electromagnetic field calculation results in each iteration process. After the number of iterations is reached, the electromagnetic field calculation results obtained in each iteration are exported as the final time-domain electromagnetic simulation results.

[0056] The beneficial effects of the present invention are as follows: the present invention introduces the concept of projection operator at the boundary, proves that the internal energy change of the distributed sum-simultaneous approximation term discrete system is determined by the electromagnetic field component at the boundary, introduces subgrid technology, proves that the distributed sum-simultaneous approximation term algorithm still maintains system energy stability in subgrid analysis, and proves the long-term stability of the distributed sum discrete system. Therefore, this calculation method can improve the calculation efficiency and accuracy for electromagnetic field simulation problems with fine structures. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a flow chart of the method of the present invention;

[0058] Figure 2 It is a schematic diagram of two-dimensional grid division;

[0059] Figure 3 It is a schematic diagram of a block uniform grid with a grid spacing ratio of 2:1;

[0060] Figure 4 A schematic diagram of the substrate integrated waveguide calculation area setting in the embodiment;

[0061] Figure 5 Schematic diagram of the time domain electric field calculation results using the traditional finite-difference time-domain method and the distributed summation-simultaneous approximation term method for different grid divisions;

[0062] Figure 6 Schematic diagram of electromagnetic field distribution obtained by different methods. DETAILED DESCRIPTION

[0063] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0064] The present invention is based on the finite difference algorithm of Maxwell's equations based on the distributed sum method and the simultaneous approximation term technology. By introducing the concept of projection operator at the boundary, it is proved that the internal energy change of the distributed sum discrete system is determined by the electromagnetic field component at the boundary, and then the two-dimensional distributed sum finite difference is theoretically derived, and finally its long-term stability is proved, and MATLAB simulation analysis is performed. On this basis, the sub-grid technology is introduced to prove that the distributed sum algorithm still maintains the system energy stability in the sub-grid analysis, and the long-term stability of the distributed sum discrete system is proved, specifically:

[0065] like Figure 1 As shown, a time domain electromagnetic simulation method based on the distributed summation method and the simultaneous approximation term includes the following steps:

[0066] A time-domain electromagnetic simulation method based on a distributed summation method and a simultaneous approximation term, characterized in that it comprises the following steps:

[0067] S1. Define the simulation target area, discretize the main grid, and give the known information of each grid point;

[0068] S2. Sub-grid local refinement discretization: Based on the main grid, each main grid point is further divided into smaller sub-grid points;

[0069] S3. Based on the distributed summation operator that maintains the long-term stability of energy, the Maxwell equations corresponding to the space are defined and set so that the corresponding semi-discrete equations can describe the changes of the electromagnetic field in the space, and then the obtained Maxwell equations are used to calculate the electromagnetic field at each grid;

[0070] S4. Setting of boundary conditions based on simultaneous approximation terms: Based on the boundary conditions of the simultaneous approximation terms with adjusted energy, each boundary condition is set and the energy is regulated so as to achieve long-term stability, and the electromagnetic field under the boundary conditions is calculated;

[0071] S5. Select the time step and number of iterations: Set the time step length and the total simulation time. The single time interval t is determined by the subgrid size and is The number of iterations is the ratio of the total computation time T to a single time interval t;

[0072] S6. In each iteration process, steps S3 to S4 are repeatedly executed to obtain the electromagnetic field calculation results in each iteration process. After the number of iterations is reached, the electromagnetic field calculation results obtained in each iteration are exported as the final time-domain electromagnetic simulation results.

[0073] 1. Solution formula for Maxwell's equations based on the distributed summation method and simultaneous approximation term technology

[0074] Maxwell's equations in two dimensions have the following form:

[0075]

[0076] Among them, H x , H y are the magnetic field components in the x and y directions respectively, E z is the magnetic field component in the z direction, μ is the magnetic permeability of the grid material, and ε is the electrical conductivity of the grid material. The grid distribution used is consistent with the spatial grid of the traditional two-dimensional transverse magnetic wave mode (TM) finite difference time domain method, such as Figure 2 shown.

[0077] In the above rectangular domain S=(x L ,x R )×(y L ,y R ), the energy of the system is defined as:

[0078]

[0079] Taking the time derivative of the above equation and applying the divergence theorem, we get

[0080]

[0081] Among them, the vector Indicates [H x ,H y ] T , represents the unit normal vector pointing outward, Indicates the entire area. Represent the lower boundary, upper boundary, left boundary and right boundary of the region respectively. Consider the distributed summation operator of the two-dimensional region. First, each two-dimensional grid is composed of the tensor product of its corresponding one-dimensional subgrid. In addition, the column vector E z 、E z, H y To represent the spatially discretized electromagnetic field solution variables.

[0082] Next, we introduce two correlated one-dimensional distribution sum discrete matrices to construct a two-dimensional distribution sum operator. First, use a y , and Representing a one-dimensional system in the y direction:

[0083]

[0084] The domain is (y L ,y R ), and the physical quantity in this direction is discretized according to the difference grid as and matrix a y , and Defined as:

[0085]

[0086]

[0087]

[0088]

[0089] In order, define a y and Correspondingly, and Correspondingly, we can further and q y Make the following definition:

[0090]

[0091]

[0092]

[0093] Considering the properties of the discretized distribution sum matrix, q y It can be written as:

[0094]

[0095] Similarly, the corresponding definition can be made in the x direction:

[0096]

[0097] The domain is (x L ,x R ), and the matrix |a x 、a| x , and Same as the definition of the y direction. Similarly, for the matrix and q x Make the following definition:

[0098]

[0099]

[0100]

[0101] Using these one-dimensional distribution sum discretization matrices, the two-dimensional distribution sum discretization matrix can be constructed as follows:

[0102]

[0103]

[0104] Among them, the symbol represents the Kronecker product, that is, the tensor product calculation of matrices.

[0105] Through the above matrix, the two-dimensional Maxwell equations can be discretized into:

[0106]

[0107] The energy of the semi-discrete system is defined as:

[0108]

[0109] With these discrete matrices and their properties, we can perform discrete energy analysis on the system. Taking the time derivative of the above formula, we can get:

[0110]

[0111] Based on the derivation of the above semi-discrete formula, it can be proved that the energy of a continuous system can be represented by a discrete system, and the energy of a discrete system depends on its boundary field. Taking the boundary condition of an ideal electric conductor as an example, the boundary condition of an ideal electric conductor should be that the electric field at the boundary is 0. Adding the simultaneous approximation term technology to the semi-discrete system, we get:

[0112]

[0113] Among them, σ L , σ R , σ B and σ Tis an unknown scalar. In order to make the system energy stable, the rate of change of energy over time must be 0, that is:

[0114] σ L =-1σ R =1σ B =-1σ T =1

[0115] 、、、。

[0116] 2. The long-term stable distribution summation-simultaneous approximation FDTD subgrid method can be theoretically proved

[0117] The entire analysis area was z The horizontal interface of subgrid alignment is divided into two blocks, and the grid point distribution of each block area conforms to the distribution sum discrete grid distribution. The electromagnetic components E of the top block and the bottom block z The subgrids all have grid points assigned to them on the staggered grid interface, but the grid spacing is different. Figure 3 A schematic diagram of a subgrid partition with a grid spacing ratio of 2:1 is given in . To distinguish, the symbols associated with the top are indicated by adding a superscript “+”, and the symbols associated with the bottom are indicated by adding a superscript “-”.

[0118] Ignoring the interface conditions and boundary conditions of the staggered grid interface, in each block, the electromagnetic field system can be obtained using the given distributed summation operator to obtain a semi-discrete system as shown below:

[0119]

[0120]

[0121] The definitions of the symbols are the same as those in the previous text, and the energy of each region can be expressed as

[0122]

[0123]

[0124] Therefore, the total system energy E is defined as the block system energy E - and E + The sum.

[0125] For the energy of the block system, it can be obtained:

[0126]

[0127]

[0128] Among them, the one-dimensional column vector and is defined as:

[0129]

[0130] and are all vectors of equal magnitude to P, The only non-zero element in corresponds to E z - The last element of The only non-zero element in corresponds to E z + The first element of . Two-dimensional distribution sum projection operator and Both are related to H x Column vectors of equal size can be defined as

[0131]

[0132] The matrix Q represents the locations of the staggered grid interfaces:

[0133]

[0134] The matrix Q has only two non-zero rows, corresponding to E z - The last element and E z + The first element of .

[0135] Adding simultaneous approximation boundary conditions to the block system and considering the staggered grid interface conditions, we can obtain

[0136]

[0137]

[0138] in, is an unknown scalar parameter, defining a one-dimensional column vector:

[0139]

[0140]

[0141] These one-dimensional column vectors can be understood as the projections of their respective two-dimensional domains onto the interface. In general, the vector and cannot be added or subtracted from each other because the meshes on both sides of the interface are of different sizes. Therefore, it is necessary to introduce the interface transfer operation matrix and

[0142] in, It is a suitable and The difference operator, so that the resulting vector and Can be added or subtracted. The effect is similar to that of and Can be added or subtracted.

[0143] Through discrete energy analysis, we can get:

[0144]

[0145] It is known that the distribution summation property requires the boundary to be a first-order difference and the interior to be a second-order difference, and the ratio of the main grid to the subgrid is known to be n, so the interface interpolation matrix can be set In the following form:

[0146]

[0147] about and In addition to the accuracy requirements, the derivation of also imposes additional mutual conversion relationship restrictions:

[0148]

[0149] The interpolation matrix can be obtained and They are

[0150]

[0151]

[0152] Then we can get

[0153]

[0154] When the unknown scalar satisfies the above relationship, the rate of change of energy over time is 0, and the energy of the semi-discrete system is stable.

[0155] In the embodiment of the present application, a substrate integrated waveguide with two rows of ideal electrical conductor pillars on the sidewall is simulated, such as Figure 4 As shown in the figure. The radius of each pillar is 4 mm. The entire computational area is divided into 15 sub-areas, and the distributed sum-simultaneous approximation term subgrid method is used to simulate the area where the ideal electrical conductor pillar is located. In other areas, the distributed sum-simultaneous approximation term method uses a coarse grid.

[0156] The size of the coarse grid is 1 mm and the size of the subgrid is 0.5 mm. A 7.5 GHz sinusoidal current source is used as excitation. The simulation time is 8 nanoseconds. The sinusoidal source is placed at the position (17 cm, 5 cm) in the coarse grid area. For fair comparison, a time step of 1.1675 picoseconds is used for all simulations.

[0157] Figure 5 shows the comparison of the results of different methods in the time domain, where Figure 5 (a) is [0,2.650×10 -9 ]s, Figure 5 (b) is [2.260×10 -9 ,2.282×10 -9 ]s, it can be found that the results of different methods are very similar. However, the accuracy of the proposed DSS-SAP method with subgrid is similar to that of the traditional FDTD and DSS-SAP method with fine grid, and is better than that of the traditional FDTD and DSS-SAP method with coarse grid.

[0158] The electromagnetic field distributions obtained by the traditional finite-difference time-domain method and the distributed sum-simultaneous approximation term subgrid method under different grids are compared. Figure 6 As shown. Figure 6 In (a), due to the use of coarse grids in the traditional finite-difference time-domain method. Figure 6 (b), more details are shown due to the fine mesh. Figure 6 (c) shows the results of the distributed summation-simultaneous approximation subgrid method. Since a fine grid is used to cover the two regions of the ideal conductor column, Figure 6 (c) and Figure 6 There is no obvious difference in the settlement results for the ideal conductor column region in (b).

[0159] The calculation results of the substrate integrated waveguide example solved using different methods are shown in the following table:

[0160]

[0161]

[0162] Similarly, the computational time and efficiency are also compared. The following table shows the relative error, total number of elements, and computational time consumption of different methods. 6 grids, and in the fine grid there are 8×10 6 In the sub-grid, the number of grids (3.824×10 6) slightly increased. As shown in the following table, the finite-difference time-domain method and the distributed sum-simultaneous approximation method with fine grids can obtain accurate results. The calculation time of these two methods is 183.32s and 211.98s respectively. For the proposed distributed sum-simultaneous approximation subgrid method, since the fine grid is only used for a small area containing an ideal electrical conductor column, the total number of grids is 3.824×10 6 However, the relative error is only 0.57%. This time-domain electromagnetic simulation method has the advantage of analyzing fine structure electromagnetic problems in the time domain.

[0163] The above is a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in other combinations, modifications and environments, and can be modified within the scope of the concept described herein through the above teachings or the technology or knowledge of the relevant field. The changes and modifications made by those skilled in the art do not deviate from the spirit and scope of the present invention, and should be within the scope of protection of the claims attached to the present invention.

Claims

1. A time-domain electromagnetic simulation method based on the distributed summation method and simultaneous approximation term, Features: The following steps are involved: S1. Define the simulation target area, discretize the main grid, and give the known information of each grid point; S2. Sub-grid local refinement discretization: Based on the main grid, each main grid point is further divided into smaller sub-grid points; S3. Based on the distributed summation operator that maintains the long-term stability of energy, the Maxwell equations corresponding to the space are defined and set so that the corresponding semi-discrete equations can describe the changes of the electromagnetic field in the space, and then the obtained Maxwell equations are used to calculate the electromagnetic field at each grid; The step S3 comprises: S301. Given a distributed summation operator that maintains long-term energy stability, including: matrix and |a x 、a| x , and in, S302. Using these one-dimensional distribution sum discretization matrices, construct a two-dimensional distribution sum discretization matrix as follows: I x is the unit diagonal matrix corresponding to the number of grid points in the x direction of the target area, The unit diagonal matrix corresponding to the number of grid points in the y direction of the target area; S303. The two-dimensional Maxwell equations are in the following form: Among them, H x , H y are the magnetic field components in the x and y directions respectively, E z is the magnetic field component in the z direction, μ is the magnetic permeability of the mesh material, and ε is the dielectric constant of the mesh material; Based on the two-dimensional distribution sum discretization matrix, the Maxwell equations corresponding to the space are defined and set, and the two-dimensional Maxwell equations are discretized into: S304. When the target area is a main grid and a sub-grid area, according to steps S302 to S303, the corresponding two-dimensional distribution sum discretization matrix is ​​calculated respectively, and then the two-dimensional Maxwell equations of the main grid and the sub-grid area are obtained respectively, so as to calculate the electromagnetic field value of each grid point of the main grid and the sub-grid area; the electromagnetic field value calculated for each grid point includes H x , H y and E z ; S4. Setting of boundary conditions based on simultaneous approximation terms: Based on the boundary conditions of the simultaneous approximation terms with adjusted energy, each boundary condition is set and the energy is regulated so as to achieve long-term stability, and the electromagnetic field under the boundary conditions is calculated; The step S4 comprises: S401. On the premise of discretization of the two-dimensional Maxwell equations in step S3, adding boundary conditions of simultaneous approximation terms to perform energy regulation; For closed boundary conditions, after adding the boundary conditions of the simultaneous approximation term, the obtained two-dimensional Maxwell equations are: Among them, σ L , σ R , σ B and σ T is a scalar; Take σ L =-1,σ R =1,σ B =-1,σ T =1, so that under the closed boundary condition, the rate of change of the electromagnetic field energy in the region with time is always 0, thereby achieving long-term stability of energy by simultaneously approximating the boundary conditions; S402. After adding the boundary conditions of the simultaneous approximation term for the main grid and sub-grid boundaries, the obtained two-dimensional Maxwell equations are: The main grid boundaries are: The subgrid boundaries are: in, is an unknown scalar parameter; Pick All are -1 / 2, so that the electromagnetic energy of the main grid and the sub-grid interacts, and the time conversion rate is always 0, maintaining the long-term stability of energy; Among them, the interpolation matrix and They are S403. According to steps S401 to S402, the electromagnetic fields of the closed boundary, the main grid boundary and the sub-grid boundary are calculated; S5. Select the time step and number of iterations: Set the time step length and the total simulation time. The single time interval t is determined by the subgrid size and is The number of iterations is the ratio of the total computation time T to a single time interval t; S6. In each iteration process, steps S3 to S4 are repeatedly executed to obtain the electromagnetic field calculation results in each iteration process. After the number of iterations is reached, the electromagnetic field calculation results obtained in each iteration are exported as the final time-domain electromagnetic simulation results.

2. A time domain electromagnetic simulation method based on distributed summation method and simultaneous approximation term according to claim 1, Features: The step S1 comprises: S101. Determine the simulation area: Given a target object to be simulated, and set the size range of the entire rectangular simulation area containing the target object: length is x, width is y; S102. Determine the boundary condition type: The boundary condition type is a closed boundary condition, which means that when electromagnetic waves in the simulation area propagate to the boundary of the simulation area, reflection will be directly generated; S103. Main grid discretization: According to the wavelength λ of the simulated electromagnetic wave, the resolution of spatial discretization is determined, and a single main grid size of (λ / 20)×(λ / 20) square grid is set; after the grid covers the entire calculation area, discrete grid points are set in the simulation area. These grid points will be used to record the values ​​of the electric and magnetic fields. The number of grid points in the x direction is x / (λ / 20)+1, and the number of grid points in the y direction is y / (λ / 20)+1. For each grid point, the following information is included: grid point position (x ij ,y ik ), the electric field strength E at the grid point i , magnetic field strength H i ; S104 determines the material distribution: determines the materials of each part in the simulation area, describes the material parameters of each grid point, and assigns them. The assigned parameters include: the material's magnetic permeability μ, the material's dielectric constant conductivity ε, and the material's electrical conductivity σ ; S105. Setting initial conditions: At the beginning of the simulation, it is necessary to specify the initial conditions of the electric field and the magnetic field, including the distribution of the electric field and the magnetic field, that is, specifying the magnetic field and the electric field at each grid point; Specify the information of the excitation source. If point current source excitation is used, the following information is included: excitation source position (x i ,y i ), excitation amplitude U, excitation source waveform change f, excitation source wavelength λ.

3. A time domain electromagnetic simulation method based on distributed summation method and simultaneous approximation term according to claim 2, Features: In step S2, the fine structure position where the resolution needs to be increased is set as region (x s1 ,y s1 )×(x s2 ,y s2 ), the grid refinement ratio is 2 times, then the number of grid points in the sub-grid x direction is: (x s2 -x s1 ) / (λ / 20 / 2)+1, the number of grid points in the y direction is: (y s2 -y s1 ) / (λ / 20 / 2)+1, the information contained in each grid point is consistent with the main grid point.