Dielectric electromagnetic effect estimation method of variable step FETD based on Sigmoid

Abstract

The invention discloses a dielectric electromagnetic effect estimation method of variable step size FETD (Field Effect Transistor Difference) based on Sigmoid. The method comprises the following steps: constructing a simulation model of a dielectric body and an estimation region under the excitation of a preset plane electromagnetic wave, and carrying out grid division so as to establish a time domain iteration matrix of the estimation region; performing time step iteration on the time domain iteration matrix within a preset total time to obtain an error mean value of the single-step space electric field intensity of the estimation region; and updating the preset time step by using a variable step size adjustment method based on Sigmoid function improvement, stopping iteration until the preset simulation time is reached, and obtaining the electric field distribution of the estimation region of the dielectric body in the preset total time, thereby obtaining the electromagnetic effect parameter of the estimation region. The method can achieve the dynamic self-adaption of the time step length in the time domain finite element method, improves the estimation efficiency, reduces the estimation time, can more accurately control the estimation precision of the electric field, suppresses the numerical oscillation of the estimation precision, improves the stability, and has a very high practical engineering application value.

Description

Technical Field

[0001] The present invention relates to a method for estimating the electromagnetic effect of a dielectric, and particularly to a method for estimating the electromagnetic effect of a dielectric based on Sigmoid variable-step FETD. Background Art

[0002] When performing time-domain finite element simulation of electromagnetic fields to solve some structures with high accuracy requirements, there are usually two options: 1. Reduce the mesh division size, use a denser mesh to discretize the target object or increase the order of the basis function to control the solution accuracy. 2. Adopt a time-step adaptive finite element time domain (FETD) algorithm in the time domain to control the solution accuracy. However, the former method will bring more unknowns, and the selection of the time step will also be limited by the minimum division size. Moreover, for electrically large sizes, the method of using high-order basis functions takes a long time to calculate. In comparison, controlling the time step in the time domain is more stable and flexible in manipulating the solution accuracy, and can also significantly improve the calculation efficiency. The existing variable-step accelerated finite element time domain (FETD) methods mainly include the dichotomy method and the variable-step method based on exponential functions. Both can improve the solution efficiency to a certain extent, but the stability of their solution accuracy needs to be improved. Summary of the Invention

[0003] To solve the problems existing in the background art, the present invention provides a method for estimating the electromagnetic effect of a dielectric based on Sigmoid variable-step FETD. Based on the fixed-step fourth-order single diagonal implicit Runge-Kutta (SDIRK4) difference format and its dichotomy variable-step acceleration strategy, the estimation accuracy and step-size adaptive method of the Sigmoid-based variable-step strategy are studied. This method can estimate the electromagnetic field target problem more quickly, efficiently and with stable accuracy.

[0004] The technical solution adopted by the present invention is as follows:

[0005] The method for estimating the electromagnetic effect of a dielectric based on Sigmoid variable-step FETD of the present invention includes:

[0006] First, construct a simulation model of the dielectric and its estimation region under the excitation of a preset plane electromagnetic wave and perform mesh division, so as to establish a time-domain iterative matrix of the estimation region of the dielectric in the simulation model.

[0007] Second, perform time-step iteration on the time-domain iterative matrix within a preset total time. Each time a preset time step is iterated, obtain the mean error of the single-step spatial electric field strength in the estimation region.

[0008] In the third step, according to the mean error of the single-step spatial electric field intensity in the estimated region, the preset time step is updated using a variable step-size adjustment method improved based on the Sigmoid function, and the updated time step is used for the iterative estimation of the next time step until the iteration stops after reaching the preset simulation time, obtaining the electric field distribution in the estimated region of the dielectric within the preset total time, thereby obtaining the electromagnetic effect parameters of the estimated region.

[0009] In the first step described above, the estimated region of the dielectric is the outer edge region of the dielectric; the simulation model is meshed using tetrahedral meshes, and after meshing, the structural information of the simulation model is obtained, including the node information and element information after meshing. The node information includes the node number and node coordinates, and the element information includes the element number and the node numbers included in each element; the simulation parameters are set, and the structural information of the model can be read.

[0010] In the first step described above, the time-domain iteration matrix of the estimated region of the dielectric in the simulation model is as follows:

[0011]

[0012] [T] ij =∫ Ω εN i ·N j dΩ

[0013]

[0014] Among them, [T] ij 、[R] ij and [S] ij respectively represent the first, second, and third coefficient matrices obtained in the construction of the time-domain iteration matrix. i and j respectively represent the first and second dimensions of the first, second, and third coefficient matrices; E represents the electric field intensity in the estimated region of the dielectric under the excitation of the preset plane electromagnetic wave; t is the time; {ξ i} represents the externally applied electromagnetic wave excitation at the i-th grid node; ε and μ respectively represent the permittivity and permeability; N i and N j respectively represent the first-order basis function at the i-th grid node and the weighted shape function at the j-th grid node; Ω represents the overall range of the estimated region; σ represents the conductivity; Γ and Γ N respectively represent the boundary of the estimated region and the Neumann boundary part therein; Y represents the intermediate parameter; represents the normal vector of the unit grid surface; ▽ represents the Hamiltonian operator; J imp represents the preset current density excitation in the estimated region; W i represents the weighted shape function at the i-th Neumann boundary; K Nis a known function on the Neumann boundary; N D represents the total number of grid nodes on the Dirichlet boundary; represents the weighted shape function at the j-th Dirichlet boundary grid node; represents the electric field strength at the j-th Dirichlet boundary grid node; k 0 represents the wave number of the electromagnetic wave, η r represents the Neumann boundary Γ N the normalized surface impedance on it.

[0015] In the second step described above, according to the electric field strength of the preset plane electromagnetic wave excitation at the initial moment in the estimation region, the difference format of the fourth-order single diagonal implicit Runge-Kutta SDIRK4 is used for the time-domain iteration matrix within the preset total time, and the time-step iteration of the finite element in time domain FETD is performed. Two similar Butcher matrices are randomly adopted during the time-step iteration; the mean error error of the single-step spatial electric field strength of the obtained estimation region is as follows:

[0016]

[0017] where n represents the total number of nodes in the estimation region divided by tetrahedral meshes; E ia and E ib respectively represent the electric field strength of the k-th node in the estimation region at time t m when the fourth-order single diagonal implicit Runge-Kutta SDIRK4 difference format adopts two similar Butcher matrices at time t m+1 obtained at time t.

[0018] In the third step described above, the mean error of the single-step spatial electric field strength of the estimation region is used as the estimation accuracy at the current moment, and the time step is updated using a variable step size adjustment method improved based on the Sigmoid function, as follows:

[0019] step t+1 = step t ×(1 + k)

[0020]

[0021] where step t and step t+1 respectively represent the updated time steps at the current time t and the next time t + 1; k represents the step size adjustment factor improved based on the Sigmoid function; A and B respectively represent the first and second adjustment parameters; error t represents the estimation accuracy at the current time t, and error exp represents the preset estimation accuracy.

[0022] In the third step described above, the time step is adjusted according to a preset deviation threshold δ as follows:

[0023] If -δ ≤ error t -error exp ≤ δ, then the current time step is not adjusted;

[0024] If error t -error exp > δ, then the variable time step adjustment method improved based on the Sigmoid function is used to update the time step, and the time step is decreased;

[0025] If error t -error exp < -δ, then the variable time step adjustment method improved based on the Sigmoid function is used to update the time step, and the time step is increased.

[0026] In the third step described above, during the iteration process, the electric field intensity at each node in the simulation model is updated through the time domain iteration matrix of the estimated region of the dielectric until the iteration ends, and the final electric field intensity at each node is obtained, and then the electric field distribution of the estimated region of the dielectric is obtained.

[0027] According to the magnitude relationship between the current estimation accuracy and the preset estimation accuracy, the time step size for the next moment is adjusted, so as to realize the time step self - adaptation of the finite - element time - domain FETD method. Then, the electric field intensity at time t is updated by the iteration matrix equation of the fourth - order single - diagonal implicit Runge - Kutta SDIRK4 difference format, and the correct electric field intensity value can be obtained at each point in space. Thus, the process of time step update is all completed; repeat the steps until the time iteration ends. m+1 At this time, the update process of the time step is all completed; repeat the steps until the time iteration ends.

[0028] The electronic device of the present invention includes: a memory and a processor that are coupled to each other, wherein the memory stores program data, and the processor calls the program data to execute the method as described above.

[0029] The computer - readable storage medium of the present invention stores program data thereon, and when the program data is executed by a processor, the method as described above is implemented.

[0030] The beneficial effects of the present invention are:

[0031] The present invention can realize the dynamic self - adaptation of the time step in the finite - element time - domain method. Compared with the fixed - step - size algorithm of the finite - element time - domain FETD and the current variable - step - size methods such as the dichotomy method, the estimation efficiency of the finite - element time - domain method is improved, the estimation time is reduced, the estimation accuracy of the electric field can be controlled more accurately, the numerical oscillation of the estimation accuracy is suppressed, and the stability of the estimation accuracy is improved, which has strong practical engineering application value. Description of the Drawings

[0032] Figure 1 FIG. is a schematic diagram of the effect of a metal cylinder on an incident plane electromagnetic wave obtained by simulating and solving the method of the present invention;

[0033] Figure 2 FIG. is a schematic diagram of the effect of a metal cylinder on an incident plane electromagnetic wave at the same moment obtained by simulating the binary search variable step size finite element time domain (FETD) method;

[0034] Figure 3 FIG. is a time step change diagram obtained by using the method of the present invention;

[0035] Figure 4 FIG. is a schematic diagram of the step size adjustment function improved by the Sigmoid of the present invention;

[0036] Figure 5 FIG. is a comparison diagram of the estimation accuracy varying with the time step obtained by using the method of the present invention and the traditional double-length variable step size method in the estimation region. Detailed Embodiment

[0037] The present invention will be further described in detail below with reference to the drawings and specific embodiments.

[0038] The method for estimating the electromagnetic effect of a dielectric based on Sigmoid variable step size FETD of the present invention is specifically as follows:

[0039] First step, construct a simulation model of the dielectric and its estimation region under the excitation of a preset plane electromagnetic wave and perform mesh division, so as to establish a time domain iteration matrix of the estimation region of the dielectric in the simulation model; the estimation region of the dielectric is the outer edge region of the dielectric; use tetrahedral meshes to perform mesh division on the simulation model, and after mesh division, obtain the structural information of the simulation model, including node information and element information after mesh division, the node information includes node numbers and node coordinates, and the element information includes element numbers and the node numbers included in each element; set simulation parameters, and the structural information of the model can be read.

[0040] The time domain iteration matrix of the estimation region of the dielectric in the simulation model is as follows:

[0041]

[0042] [T] ij =∫ Ω εN i ·N j dΩ

[0043]

[0044]

[0045] Among them, [T] ij , [R] ij and [S] ij respectively represent the first, second, and third coefficient matrices obtained in the construction of the time-domain iterative matrix. i and j respectively represent the first and second dimensional lengths of the first, second, and third coefficient matrices; E represents the electric field strength in the estimated region of the dielectric under the excitation of a preset plane electromagnetic wave; t is the time; {ξ i} represents the externally applied electromagnetic wave excitation at the i-th grid node; ε and μ respectively represent the permittivity and permeability; N i and N j respectively represent the first-order basis function at the i-th grid node and the weighted shape function at the j-th grid node; Ω represents the overall range of the estimated region; σ represents the conductivity; Γ and Γ N respectively represent the boundary of the estimated region and the Neumann boundary part therein; Y represents an intermediate parameter; represents the normal vector of the unit grid surface; represents the Hamiltonian operator; J imp represents the preset current density excitation in the estimated region; W i represents the weighted shape function at the i-th Neumann boundary; K N is a known function on the Neumann boundary; N D represents the total number of grid nodes on the Dirichlet boundary; represents the weighted shape function at the j-th Dirichlet boundary grid node; represents the electric field strength at the j-th Dirichlet boundary grid node; k 0 represents the wave number of the electromagnetic wave, η r represents the normalized surface impedance on the Neumann boundary Γ N .

[0046] When establishing the time-domain iterative matrix, first, a wave equation containing only the electric field vector is derived based on Maxwell's equations; then, according to the known plane electromagnetic wave excitation set in the estimated region, both sides of the equation are tested using the Galerkin method and expanded with basis functions to obtain the final time-domain iterative formula, and the time-domain iterative matrix is filled according to the iterative formula.

[0047] Construct the first two equations of the time-domain form of Maxwell's equations with the electric field strength E and magnetic field strength H as unknowns in the estimated region, as follows:

[0048]

[0049] Eliminate the magnetic field strength H from the above formula to obtain a wave equation containing only the electric field vector. Perform a Galerkin test on both sides of the equation and then expand it with basis functions to obtain:

[0050]

[0051] Perform the Galerkin test on the equation. Expand the electric field strength E and the magnetic field strength H using basis functions as follows:

[0052] Eliminate the magnetic field from the first two equations of the Maxwell's equations in time domain form to obtain the wave equation containing only the electric field vector as follows:

[0053]

[0054] Substitute the typical homogeneous Dirichlet boundary condition and the mixed boundary condition of the impedance surface:

[0055] On Γ D above

[0056] On Γ N above

[0057] where P represents the tangential electric field value on the Dirichlet boundary part Γ D above; μ r represents the relative magnetic permeability.

[0058] Substitute the formula and perform the Galerkin test on the wave equation containing the Dirichlet boundary condition and the mixed boundary condition. Through vector identities and divergence theorem, the weak solution expression of this boundary value problem is obtained as follows:

[0059]

[0060] Expand the electric field vector E using basis functions. Select the basis functions in the expansion formula to have the same form as the weighted shape functions, and organize to obtain the time domain iteration matrix.

[0061] When filling [T], [R], [S] and {ξ}, fill them in sequence according to the encoding of the tetrahedral elements. Since each element has no interaction with other elements, the obtained matrix [T] has the block diagonal property.

[0062] In the second step, perform time step iteration on the time domain iteration matrix within the preset total time. For each iteration of the preset time step, obtain the mean error of the single-step spatial electric field strength in the estimation region; According to the electric field strength of the preset plane electromagnetic wave excitation at the initial moment in the estimation region, use the fourth-order single diagonal implicit Runge-Kutta SDIRK4 difference format to perform time step iteration of the finite element in time domain FETD on the time domain iteration matrix within the preset total time. Randomly adopt two similar Butcher matrices during the time step iteration; The mean error error of the single-step spatial electric field strength in the obtained estimation region is as follows:

[0063]

[0064] Among them, n represents the total number of nodes in the estimated region of tetrahedral mesh generation; E ia and E ib respectively represent the electric field strengths of the k-th node in the estimated region at time t m when the fourth-order singly diagonal implicit Runge-Kutta SDIRK4 difference scheme adopts two similar Butcher matrices at time t m+1 in the estimated region at time t

[0065] The matrix equation is numerically approximated according to the difference scheme of the fourth-order singly diagonal implicit Runge-Kutta SDIRK4 to obtain:

[0066]

[0067] Among them, E t+coΔt 、E t+cqΔt and E t+1 respectively represent the estimated electric field strengths within a single time step at times t + c o Δt, t + c q Δt, and t + 1 when the difference scheme of the fourth-order singly diagonal implicit Runge-Kutta SDIRK4 is iterated; c o and c q respectively represent the elements of the o-th and q-th rows of the c vector in the Butcher matrix; Δt represents the single-step time step size; represents the initial electric field strength at time t; a oq represents the element in the o-th row and q-th column of the a matrix in the Butcher matrix; I n represents the n-order identity matrix; f() represents an intermediate variable; b q represents the element of the q-th column of the b vector in the Butcher matrix; T represents the T matrix in the time-domain iterative matrix equation; ξ t+ciΔt represents the externally applied electromagnetic wave excitation at time t + c o Δt within the time step; S and R respectively represent the S matrix and R matrix in the time-domain iterative matrix equation; represents the derivative of the electric field strength E.

[0068] Within the estimated region, the mean value of the single-step error of the electric field strength calculated by the variable time step strategy at time t m+1 under the fourth-order singly diagonal implicit Runge-Kutta SDIRK4 difference scheme with two pre-given different estimated Butcher matrices.

[0069] Examples of the selection of the two pre-set estimated Butcher matrices are as follows:

[0070]

[0071]

[0072] In the third step, according to the mean error of the single-step spatial electric field intensity in the estimated region, use the variable step-size adjustment method improved based on the Sigmoid function to update the preset time step, and use the updated time step to perform iterative estimation for the next time step until the iteration stops after reaching the preset simulation time, and obtain the electric field distribution of the estimated region of the dielectric within the preset total time, so as to obtain the electromagnetic effect parameters of the estimated region.

[0073] Take the mean error of the single-step spatial electric field intensity in the estimated region as the estimation accuracy at the current moment, and use the variable step-size adjustment method improved based on the Sigmoid function to update the time step as follows:

[0074] step t+1 = step t ×(1 + k)

[0075]

[0076] where step t and step t+1 represent the updated time steps at the current t moment and the next t + 1 moment respectively; k represents the step-size adjustment factor improved based on the Sigmoid function; A and B represent the first and second adjustment parameters respectively; error t represents the estimation accuracy at the current t moment, and error exp represents the preset estimation accuracy.

[0077] The first parameter A of the step-size adjustment factor is used to control the maximum adjustment range of the step size for the next time step, and the second parameter B is used to adjust the stretching of the Sigmoid improved function, thereby controlling the sensitivity of the step-size adjustment amplitude to the deviation δ between the current estimation accuracy and the preset numerical accuracy.

[0078] The adjustment of the time step size is adjusted according to the preset deviation threshold δ as follows:

[0079] If -δ ≤ error t - error exp ≤ δ, then do not adjust the current time step.

[0080] If error t - error exp > δ, then use the variable step-size adjustment method improved based on the Sigmoid function to update the time step and decrease the time step.

[0081] If error t - error expIf < -δ, the time step is updated using the variable step size adjustment method improved based on the Sigmoid function, and the time step is increased.

[0082] During the iteration process, the electric field intensity at each node in the simulation model is updated through the time-domain iteration matrix of the estimated region of the dielectric until the iteration ends, and the final electric field intensity at each node is obtained. Furthermore, the electric field distribution of the estimated region of the dielectric is obtained.

[0083] According to the magnitude relationship between the current estimation accuracy and the preset estimation accuracy, the size of the time step at the next moment is adjusted, thereby realizing the time step adaptivity of the finite element time domain (FETD) method. Then, the electric field intensity at time t is updated by the iteration matrix equation of the fourth-order single diagonal implicit Runge-Kutta (SDIRK4) difference format. m+1 The correct value of the electric field intensity can be obtained at each point in space, and the process of time step update is completed. Repeat the steps until the time iteration ends.

[0084] The electromagnetic effect parameters such as scattering of the estimated region are obtained using the obtained electric field distribution of the estimated region. Similar to the post-processing in the finite element time domain (FETD) method, it will not be elaborated here.

[0085] To verify the correctness and effectiveness of the present invention, the scattering effect of a metal cylinder on a uniformly incident plane electromagnetic wave is analyzed below. The variable step size finite element time domain (FETD) method improved based on the Sigmoid function and the fixed step size finite element time domain (FETD) method are used to calculate the change of the spatial electric field intensity in this process with the same numerical accuracy.

[0086] The calculation example is a metal cylinder with a radius of 0.25 m. The origin of the coordinate is fixed at the center of the circle, and a plane rectangular coordinate system is established. The plane electromagnetic wave is incident along the positive X-axis direction, the electric field intensity vector is linearly polarized in the Z direction, and the electromagnetic wave frequency is 300 MHz. The electric field intensity distribution at a certain moment in the solution domain when the plane wave passes through the metal cylinder solved by the variable step size FETD method improved based on the Sigmoid function is as Figure 1 shown, and the electric field intensity in the solution domain at the same moment estimated by the bisection variable step size FETD method is as Figure 2 , and the change of the time step size estimated by the variable step size FETD method improved based on the Sigmoid function for this problem is as Figure 3 shown. It can be seen that when estimated using this method, the change of the step size can quickly correspond to the preset estimation accuracy and tend to be stable. In this embodiment, the mean value of the single-step error of the spatial electric field intensity at the next step under two different preset estimated Butcher matrix configurations of the fourth-order single diagonal implicit Runge-Kutta (SDIRK4) difference format is calculated every 6 steps in the time step iteration and every 4 steps thereafter; in this embodiment, the step size of one iteration is adjusted based on the improved Sigmoid function, and the improved Sigmoid function used to adjust the step size is asFigure 4 As shown, A is the first adjustment parameter, which is used to set the change range of the coefficient multiplied in the next time step. The required estimation accuracy δ of the electric field strength preset in this embodiment is 2×10 -22 , the change of the estimation accuracy of the step-size adaptive FETD method with the number of iteration steps and the change of the estimation accuracy of the dichotomy variable-step FETD method for comparison with the number of iteration steps are as Figure 5 shown. It can be seen that, compared with the traditional variable-step estimation method using the dichotomy method, the estimation accuracy and stability of this method are higher, and the estimation accuracy can be stably estimated more quickly.

[0087] In addition, for those of ordinary skill in the art, the steps in the embodiments described in the present invention, whether all or part of them, can be implemented by programming instructions for corresponding hardware devices. The corresponding control program can be stored in various computer-readable media, such as but not limited to read-only memory, hard disk or optical disc and other media.

[0088] The above-given embodiments are only examples for implementing the present invention, and the present invention is not limited to the above embodiments. Any non-essential addition or replacement made by those skilled in the art according to the technical features of the technical solution of the present invention falls within the protection scope of the present invention.

Claims

1. A method for estimating the electromagnetic effect of a dielectric body based on a Sigmoid variable step size FETD, characterized in that: include: The first step is to construct a simulation model of the dielectric and its estimated area under the excitation of a preset plane electromagnetic wave and perform meshing, thereby establishing a time domain iteration matrix of the estimated area of ​​the dielectric in the simulation model; The second step is to iterate the time domain iteration matrix in a preset total time, and obtain the error mean of the single-step spatial electric field intensity of the estimation area at each preset time step. The third step is to update the preset time step according to the error mean of the single-step spatial electric field strength in the estimated area using a variable step adjustment method improved based on the Sigmoid function, and use the updated time step to iterate and estimate the next time step until the preset simulation time is reached and the iteration is stopped to obtain the electric field distribution of the estimated area of ​​the dielectric within the preset total time, thereby obtaining the electromagnetic effect parameters of the estimated area.

2. The method for estimating the electromagnetic effect of a dielectric body based on a Sigmoid variable step size FETD according to claim 1, characterized in that: In the first step, the estimated area of ​​the dielectric is the outer edge area of ​​the dielectric; the simulation model is meshed using a tetrahedral mesh, and structural information of the simulation model is obtained after meshing, including node information and unit information after meshing, the node information includes node numbers and node coordinates, and the unit information includes unit numbers and node numbers contained in each unit.

3. The method for estimating the electromagnetic effect of a dielectric body based on a Sigmoid variable step size FETD according to claim 2, characterized in that: In the first step, the time domain iteration matrix of the estimated area of ​​the dielectric in the simulation model is as follows: [T] ij =∫ Ω εN i ·N j dΩ Among them, [T] ij , [R] ij and [S] ij They represent the first, second and third coefficient matrices obtained in the time domain iterative matrix construction, respectively; i and j represent the first and second dimension lengths of the first, second and third coefficient matrices, respectively; E represents the electric field intensity in the estimated area of ​​the dielectric body under the preset plane electromagnetic wave excitation; t is the time; {ξ i } represents the external electromagnetic wave excitation on the i-th grid node; ε and μ represent the dielectric constant and magnetic permeability respectively; N i and N j denote the first-order basis function on the i-th grid node and the weighted shape function on the j-th grid node, respectively; Ω denotes the overall range of the estimation area; σ denotes the conductivity; Γ and Γ N They represent the boundary of the estimated region and the Neumann boundary part thereof respectively; Y represents the intermediate parameter; represents the normal vector of the unit grid surface; ▽ represents the Hamiltonian operator; J imp represents the preset current density excitation in the estimation area; W i represents the weighted shape function at the i-th Neumann boundary; K N is a known function on the Neumann boundary; N D Represents the total number of grid nodes on the Dirichlet boundary; represents the weighted shape function at the jth Dirichlet boundary grid node; represents the electric field intensity at the jth Dirichlet boundary grid node; k0 represents the wave number of the electromagnetic wave, η r represents the Neumann boundary Γ N Normalized surface impedance on .

4. The method for estimating the electromagnetic effect of a dielectric body based on a Sigmoid variable step size FETD according to claim 2, characterized in that: In the second step, according to the electric field strength of the preset plane electromagnetic wave excitation in the estimation area at the initial moment, the time domain iteration matrix is ​​subjected to the time step iteration of the time domain finite element FETD using the fourth-order single diagonal implicit Runge-Kutta SDIRK4 difference format within the preset total time, and two similar Butcher matrices are randomly used during the time step iteration; the error mean error of the single-step spatial electric field strength of the estimation area is obtained as follows: Where n represents the total number of nodes in the estimated area divided by the tetrahedral mesh; E ia and E ib Respectively represent the estimated area at t m The fourth-order single diagonal implicit Runge-Kutta SDIRK4 difference format at the time is obtained by using two similar Butcher matrices. m+1 The electric field strength of the kth node in the estimation area at time.

5. The method for estimating the electromagnetic effect of dielectrics based on Sigmoid variable step-size FETD according to claim 1, characterized in that: In the third step, the error mean of the single-step spatial electric field strength in the estimated area is used as the estimation accuracy at the current moment, and the time step is updated using the variable step adjustment method improved based on the Sigmoid function, as follows: step t+1 =step t ×(1+k) Among them, step t and step t+1 They represent the updated time steps at the current time t and the next time t+1 respectively; k represents the step adjustment factor; A and B represent the first and second adjustment parameters respectively; error t Indicates the estimation accuracy at the current time t, error exp Indicates the preset estimation accuracy.

6. The method for estimating the electromagnetic effect of a dielectric body based on a Sigmoid variable step size FETD according to claim 5, characterized in that: In the third step, the time step is adjusted according to the preset deviation threshold δ, as follows: If -δ≤error t -error exp ≤δ, the current time step will not be adjusted; If error t -error exp >δ, the variable step size adjustment method based on the Sigmoid function is used to update the time step and reduce the time step; If error t -error exp <-δ, the time step is updated using the variable step adjustment method based on the Sigmoid function to increase the time step.

7. The method for estimating the electromagnetic effect of a dielectric body based on a Sigmoid variable step size FETD according to claim 2, characterized in that: In the third step, during the iteration process, the electric field strength at each node in the simulation model is updated by the time domain iterative matrix of the estimated area of ​​the dielectric until the iteration is completed, and the final electric field strength at each node is obtained, thereby obtaining the electric field distribution of the estimated area of ​​the dielectric.

8. An electronic device, characterized in that: include: A memory and a processor coupled to each other, wherein the memory stores program data, and the processor calls the program data to execute the method according to any one of claims 1 to 7.

9. A computer-readable storage medium having program data stored thereon, characterized in that: When the program data is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images