A simulation method for a horizontally polarized high-altitude nuclear electromagnetic pulse simulator

Abstract

The invention relates to the technical field of electromagnetic pulse simulation, and discloses a simulation method for a horizontally polarized high-altitude nuclear electromagnetic pulse simulator, comprising the following steps: constructing a three-dimensional nuclear electromagnetic pulse simulation model, placing a field probe at a preset measurement point, setting an excitation source and boundary conditions, and running a full-wave simulation; collecting graph structure data when running the full-wave simulation, the graph structure data comprising nodes and edges connecting the nodes, encoding node features of the nodes in the graph structure data into sequence data, inputting the graph structure data and the sequence unit into a process model, the process model comprising a first hidden layer, a second hidden layer and an output layer; inputting the graph structure data and the sequence unit into the first hidden layer, outputting a first hidden state of each sequence unit, inputting the first hidden state of the sequence unit into the second hidden layer, outputting a second hidden state, inputting the second hidden state into the output layer, and outputting an estimated value of an electric field waveform of a point in a simulation area.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and more specifically, to a simulation method of a horizontally polarized high-altitude nuclear electromagnetic pulse simulator. Background Art

[0002] Key equipment on military aircraft, such as radar, navigation systems and communication equipment, may be exposed to electromagnetic pulse (HEMP) when performing missions. HEMP has extremely strong instantaneous electromagnetic radiation, which can cause serious interference and damage to electronic systems without proper protection, affecting the normal operation of the aircraft.

[0003] To ensure the survivability of these critical devices in a HEMP environment, a HEMP effect test is required. The test is designed to simulate the actual HEMP field, expose the equipment under test in a controlled environment, and verify its immunity and protection effect; HEMP simulation tests are usually carried out in a certain area, using an extremely broadband high-power pulse source to generate an electromagnetic radiation field equivalent to an actual HEMP. In order to fully evaluate the protection performance of the equipment under test, it is necessary to estimate the electric field waveform at any point in the test area.

[0004] Accurately estimating the electric field distribution in the test area is crucial for designing and arranging the test plan. It ensures that the equipment under test can be exposed to the HEMP electromagnetic radiation field equivalent to the real environment, so as to reliably evaluate the survivability of the protection measures and systems. Summary of the invention

[0005] The present invention provides a simulation method for a horizontally polarized high-altitude nuclear electromagnetic pulse simulator. First, a three-dimensional electromagnetic pulse simulation model is constructed in the CST (Computer Simulation Technology) software tool. When running the simulation, graph structure data is collected. With the powerful modeling ability of the graph neural network, the propagation law of electromagnetic waves in time and space can be well learned. Finally, the constructed graph structure data is encoded into sequence data and input into the model for electric field estimation, thereby improving the estimation accuracy of the electric field waveform at any spatial point and solving the problem of difficulty in estimating the electric field waveform of points in the changing scenes in the simulation area.

[0006] The present invention provides a method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator, comprising:

[0007] Step 100, according to the actual test scenario, a three-dimensional electromagnetic pulse simulation model is constructed in the CST software tool, field probes are placed at preset measurement points, excitation sources and boundary conditions are set, and full-wave simulation is run;

[0008] Step 200, collecting graph structure data when running full-wave simulation, the graph structure data including nodes and edges connecting the nodes, a node represents a measurement space point or a non-measurement space point, an edge represents the relationship between the measurement space point and the non-measurement space point;

[0009] There are edges between nodes representing measurement space points and all nodes representing non-measurement space points;

[0010] Both nodes represent non-measurement space points. If there are no other space points on the straight line connecting the two non-measurement space points, then there is an edge between the two nodes, otherwise there is no edge.

[0011] The node characteristics representing the measurement space points are equal to the electric field characteristics at the measurement space points, and the node characteristics representing the non-measurement space points are equal to the electric field characteristics at the non-measurement space points;

[0012] The electric field characteristics at the measurement space point are obtained by the electric field waveform data obtained by the field probe at the measurement space point and the three-dimensional coordinate encoding of the measurement space point, and the electric field characteristics at the non-measurement space point are obtained by the three-dimensional coordinate encoding of the non-measurement space point;

[0013] Step 300, encoding node features of nodes in the graph structure data into sequence data, the sequence data includes n sequence units, and the t-th sequence unit represents the node features of all nodes at the t-th moment;

[0014] Step 400, input graph structure data and sequence units into the first hidden layer, each time graph structure data and a sequence unit are input, and the first hidden state of each sequence unit is output accordingly; the first hidden state of the sequence unit is input into the second hidden layer, and the second hidden state is output; the second hidden state is input into the output layer, and an estimated value of the electric field waveform at any point in the simulation area is output.

[0015] In a preferred embodiment, the excitation source waveform adopts the HEMP double exponential waveform specified in the IEC61000-2-9 standard, and the polarization mode of the excitation signal is horizontal polarization; the surrounding simulated environment is air, and the boundary condition of the simulator is set to an open boundary.

[0016] In a preferred embodiment, the three-dimensional coordinates are obtained by one-hot encoding to obtain a three-dimensional coordinate vector;

[0017] The electric field waveform data is a vector containing the electric field strength value at each moment, which is spliced ​​with the three-dimensional coordinate vector of the measurement space point to obtain the electric field characteristics at the measurement space point;

[0018] The three-dimensional coordinate vector of the non-measurement space point is concatenated with a zero vector to obtain the electric field characteristics at the non-measurement space point, and the dimension of the zero vector is the same as the dimension of the vector of the electric field waveform data.

[0019] In a preferred embodiment, the calculation formula of the first hidden layer is as follows:

[0020] ;

[0021] in represents the node hidden state of the vth node, is the set of nodes that have edges with node v, is the weight matrix of the first hidden layer, is the bias term of the first hidden layer, represents the node feature of the u-th node, is a normalizing constant, It is a sigmoid function;

[0022] The node feature of a node represents the electric field feature at the measurement space point or the electric field feature at the non-measurement space point;

[0023] ;

[0024] in represents the first hidden state of the output, represents the node hidden state of the vth node, It means to concatenate the hidden states of all nodes in the graph structure data, N represents the set of all nodes in the graph structure data, is the weight parameter, is the bias parameter, Represents a sigmoid function.

[0025] In a preferred embodiment, the calculation formula of the second hidden layer is as follows:

[0026] ;

[0027] ;

[0028] ;

[0029] ;

[0030] in, , , represents the weight parameter, , , represents the bias parameter, represents the dot product, where represents the first hidden state output when the tth sequence unit inputs the first hidden layer, and They represent the t-th and t-1-th second hidden states, respectively, n≥t≥1, n represents the total number of sequence units of the sequence data, and when t=1 , tanh is the hyperbolic tangent function, Represents a sigmoid function.

[0031] In a preferred embodiment, the calculation formula of the output layer is as follows:

[0032] ;

[0033] in represents an estimation matrix of an electric field waveform, a row vector of the electric field waveform estimation matrix represents an electric field waveform vector of a non-measurement point, and the g-th component of the electric field waveform vector represents the electric field strength value at the g-th moment; represents the t-th second hidden state, Indicates the concatenation of the 1st to the nth second hidden states, n represents the total number of sequence units of the sequence data, is the weight parameter, is the bias parameter, Represents a sigmoid function.

[0034] In a preferred embodiment, the first hidden layer is pre-trained, and during the pre-training, it is connected to the pre-trained output layer, the pre-trained output layer inputs the node features of two nodes, and outputs the transfer function between the spatial points represented by the two nodes;

[0035] The calculation formula of the pre-trained output layer is as follows:

[0036] ;

[0037] in, It represents the transfer function between nodes i and j. , represents the concatenation of the hidden state of the i-th node and the hidden state of the j-th node, is the weight parameter of the output layer, is the bias parameter of the output layer, represents the S-type function, Represents the transpose of a matrix.

[0038] In a preferred embodiment, the true value calculation formula of the loss function of the first hidden layer pre-training is as follows:

[0039] ;

[0040] in, represents the transfer function between node i and node j, represents the angular frequency, and Respectively represent the electric field time domain waveforms corresponding to node i and node j, represents Fourier transform.

[0041] In a preferred embodiment, a horizontally polarized high-altitude nuclear electromagnetic pulse simulator simulation system comprises:

[0042] Model building module, used to build 3D electromagnetic pulse simulation models in CST simulation software tools;

[0043] A data acquisition module, used to collect graph structure data when running full-wave simulation;

[0044] A data encoding module, used to encode graph structure data into sequence data;

[0045] The data processing module is used to input the sequence data into the model and output the estimated value of the electric field waveform at any point in the simulation area.

[0046] The beneficial effects of the present invention are:

[0047] 1. A hybrid network combining simulation and graph neural networks with sequence models can more accurately capture the complex dynamic characteristics of time-varying electric field waveforms in large electromagnetic pulse test areas. This comprehensive use of full-wave simulation and deep learning methods makes the estimated results more consistent with the actual situation, which helps guide electromagnetic pulse protection and equipment design.

[0048] 2. By updating the predicted value of the model output, this method is adaptive and real-time, and can update the model output as the measured data changes, thereby estimating the electric field waveform at any point. This correction mechanism enables the model to respond more flexibly to different test environments and data changes, and has stronger adaptability and generalization capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a flow chart of a simulation method of a horizontally polarized high-altitude nuclear electromagnetic pulse simulator of the present invention;

[0050] Figure 2 This is the electromagnetic simulation model diagram of the horizontally polarized wire grid electromagnetic pulse simulator;

[0051] Figure 3 This is the electric field intensity distribution diagram when t=35ns and planeAtY=0m; Figure 4 This is the electric field intensity distribution diagram when t=35ns and planeAtX=0m; Figure 5 The electric field intensity distribution is 1 when t=35ns and planeAtZ=0.6m, 1m, 2m, 3m; Figure 6The second distribution of electric field strength when t=35ns and planeAtZ=0.6m, 1m, 2m, 3m; Figure 7 The electric field intensity distribution is three when t=35ns and planeAtZ=0.6m, 1m, 2m, 3m; Figure 8 The electric field intensity distribution when t=35ns and planeAtZ=0.6m, 1m, 2m, 3m; DETAILED DESCRIPTION

[0052] The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the present specification. Various examples may omit, replace, or add various processes or components as needed. In addition, the features described in some examples may also be combined in other examples.

[0053] In one embodiment of the present invention, a method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator is provided. Figure 1 As shown, including:

[0054] Step 100, according to the actual test scenario, build a three-dimensional electromagnetic pulse simulation model in the CST software tool, place field probes at preset measurement points, set the excitation source and boundary conditions, and run a full-wave simulation.

[0055] In one embodiment of the present invention, the excitation source waveform in step 100 adopts the HEMP double exponential waveform specified in the IEC61000-2-9 standard, and the polarization mode of the excitation signal is horizontal polarization; the surrounding simulated environment is air, and the boundary condition of the simulator is set to an open boundary;

[0056] Step 200, collecting graph structure data when running full-wave simulation, the graph structure data including nodes and edges connecting the nodes, a node represents a measurement space point or a non-measurement space point, an edge represents the relationship between the measurement space point and the non-measurement space point;

[0057] There are edges between nodes representing measurement space points and all nodes representing non-measurement space points;

[0058] Both nodes represent non-measurement space points. If there are no other space points on the straight line connecting the two non-measurement space points, then there is an edge between the two nodes, otherwise there is no edge.

[0059] The node characteristics representing the measurement space points are equal to the electric field characteristics at the measurement space points, and the node characteristics representing the non-measurement space points are equal to the electric field characteristics at the non-measurement space points;

[0060] The electric field characteristics at the measurement space point are obtained by the electric field waveform data obtained by the field probe at the measurement space point and the three-dimensional coordinate encoding of the measurement space point, and the electric field characteristics at the non-measurement space point are obtained by the three-dimensional coordinate encoding of the non-measurement space point;

[0061] In one embodiment of the present invention, the three-dimensional coordinates are obtained by one-hot encoding to obtain a three-dimensional coordinate vector;

[0062] The electric field waveform data is a vector containing the electric field strength value at each moment, which is spliced ​​with the three-dimensional coordinate vector of the measurement space point to obtain the electric field characteristics at the measurement space point;

[0063] The three-dimensional coordinate vector of the non-measurement space point is concatenated with a zero vector to obtain the electric field characteristics at the non-measurement space point, and the dimension of the zero vector is the same as the dimension of the vector of the electric field waveform data.

[0064] Step 300, encoding node features of nodes in the graph structure data into sequence data, the sequence data includes n sequence units, and the t-th sequence unit represents the node features of all nodes at the t-th moment;

[0065] Step 400: Input the graph structure data and the sequence unit into the first hidden layer. Each time the graph structure data and a sequence unit are input, the first hidden state of each sequence unit is output accordingly. Input the first hidden state of the sequence unit into the second hidden layer, and output the second hidden state. The output layer inputs the second hidden state, and outputs the estimated value of the electric field waveform at any point in the simulation area.

[0066] In one embodiment of the present invention, the calculation formula of the first hidden layer is as follows:

[0067] ;

[0068] in represents the node hidden state of the vth node, is the set of nodes that have edges with node v, is the weight matrix of the first hidden layer, is the bias term of the first hidden layer, represents the node feature of the u-th node, is a normalizing constant, It is a sigmoid function;

[0069] The node feature of a node represents the electric field feature at the measurement space point or the electric field feature at the non-measurement space point;

[0070] ;

[0071] in represents the first hidden state of the output, represents the node hidden state of the vth node, It means to concatenate the hidden states of all nodes in the graph structure data, N represents the set of all nodes in the graph structure data, is the weight parameter, is the bias parameter, Represents a sigmoid function.

[0072] In one embodiment of the present invention, the calculation formula of the second hidden layer is as follows:

[0073] ;

[0074] ;

[0075] ;

[0076] ;

[0077] in, , , represents the weight parameter, , , represents the bias parameter, represents the dot product, where represents the first hidden state output when the tth sequence unit inputs the first hidden layer, and They represent the t-th and t-1-th second hidden states, respectively, n≥t≥1, n represents the total number of sequence units of the sequence data, and when t=1 , tanh is the hyperbolic tangent function, Represents a sigmoid function.

[0078] In one embodiment of the present invention, the calculation formula of the output layer is as follows:

[0079] ;

[0080] in represents an estimation matrix of an electric field waveform, a row vector of the electric field waveform estimation matrix represents an electric field waveform vector of a non-measurement point, and the g-th component of the electric field waveform vector represents the electric field strength value at the g-th moment; represents the t-th second hidden state, Indicates the concatenation of the 1st to the nth second hidden states, n represents the total number of sequence units of the sequence data, is the weight parameter, is the bias parameter, Represents a sigmoid function.

[0081] In one embodiment of the present invention, the first hidden layer is pre-trained, and during the pre-training, it is connected to the pre-training output layer, which inputs the node features of two nodes and outputs the transfer function between the spatial points represented by the two nodes.

[0082] The calculation formula of the pre-trained output layer is as follows:

[0083] ;

[0084] in, It represents the transfer function between nodes i and j. , represents the concatenation of the hidden state of the i-th node and the hidden state of the j-th node, is the weight parameter of the output layer, is the bias parameter of the output layer, represents the S-type function, Represents the transpose of a matrix.

[0085] In one embodiment of the present invention, the true value calculation formula of the loss function of the first hidden layer pre-training is as follows:

[0086] ;

[0087] in, represents the transfer function between node i and node j, represents the angular frequency, and Respectively represent the electric field time domain waveforms corresponding to node i and node j, represents Fourier transform.

[0088] The present invention utilizes electromagnetic simulation software CST to simulate a large-scale horizontally polarized wire grid electromagnetic pulse simulator to be constructed, and provides simulation results of the spatial radiation field in its test area.

[0089] The proposed large-scale horizontally polarized wire grid high-altitude nuclear electromagnetic pulse simulator is used as the simulation object to analyze and calculate the spatial field strength distribution in its test area. At present, the high-altitude nuclear electromagnetic pulse environment is defined as greater than or equal to 50kV / m, and IEC61000-2-9 "Electromagnetic Compatibility (EMC) Part 2: Environment Section 9: HEMP Environment Description Radiated Disturbance" and GJB1389B "System Electromagnetic Environment Effect Requirements" have made clear provisions for it;

[0090] Figure 2This is the electromagnetic simulation model diagram of the horizontally polarized wire grid electromagnetic pulse simulator. The structure includes a biconical radiation antenna, a transmission antenna, and an absorption load resistor. The total length of the biconical antenna is 4m, the radius of the cone bottom is 1.25m, the semi-conical angle of the cone is 32°, the height of the biconical center from the ground is 10m, the distance between the two wire grid plates on the ground is 32m, the width of the single-side plate is 48m, and the transmission line width-to-height ratio is 3:2. The corresponding characteristic impedance is about 150Ω, and the single terminal load resistance is set to 600Ω. The left and right plates of the actual electromagnetic pulse simulator are multiple metal wire grids. When simulating with CST, it should be established as a solid model. The 256 transmission antennas are divided into 16 antennas as a bundle and evenly pulled out and connected to the load. There are a total of 16 integrated transmission antennas, and 8 integrated antennas are set on each side. In this example, the resistance function in the lumped element is used to set the absorption resistance value to 150Ω to realize the distributed load function. The simulator material is set to an ideal conductor, the background material is set to air, the influence of the earth medium is considered, the boundary of the calculation area is set to free space, and the maximum calculation frequency is set to 300MHz;

[0091] Figure 3 This is the electric field intensity distribution diagram when t=35ns and planeAtY=0m; Figure 4 This is the electric field intensity distribution diagram when t=35ns and planeAtX=0m; Figure 5 The electric field intensity distribution is 1 when t=35ns and planeAtZ=0.6m, 1m, 2m, 3m; Figure 6 The second distribution of electric field strength when t=35ns and planeAtZ=0.6m, 1m, 2m, 3m; Figure 7 The electric field intensity distribution is three when t=35ns and planeAtZ=0.6m, 1m, 2m, 3m; Figure 8 The electric field intensity distribution when t=35ns and planeAtZ=0.6m, 1m, 2m, 3m;

[0092] The present invention provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned horizontally polarized high-altitude nuclear electromagnetic pulse simulator simulation method is realized.

[0093] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.

Claims

1. A method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator, characterized in that: The following steps are involved: Step 100, according to the actual test scenario, a three-dimensional electromagnetic pulse simulation model is constructed in the CST software tool, field probes are placed at preset measurement points, excitation sources and boundary conditions are set, and full-wave simulation is run; Step 200, collecting graph structure data when running full-wave simulation, the graph structure data including nodes and edges connecting the nodes, a node represents a measurement space point or a non-measurement space point, an edge represents the relationship between the measurement space point and the non-measurement space point; There are edges between nodes representing measurement space points and all nodes representing non-measurement space points; Both nodes represent non-measurement space points. If there are no other space points on the straight line connecting the two non-measurement space points, then there is an edge between the two nodes, otherwise there is no edge. The node characteristics representing the measurement space points are equal to the electric field characteristics at the measurement space points, and the node characteristics representing the non-measurement space points are equal to the electric field characteristics at the non-measurement space points; The electric field characteristics at the measurement space point are obtained by the electric field waveform data obtained by the field probe at the measurement space point and the three-dimensional coordinate encoding of the measurement space point, and the electric field characteristics at the non-measurement space point are obtained by the three-dimensional coordinate encoding of the non-measurement space point; Step 300, encoding node features of nodes in the graph structure data into sequence data, the sequence data includes n sequence units, and the t-th sequence unit represents the node features of all nodes at the t-th moment; Step 400, inputting graph structure data and sequence units into the first hidden layer, each time inputting graph structure data and a sequence unit, and correspondingly outputting the first hidden state of each sequence unit; The first hidden state of the sequence unit is input into the second hidden layer, and the second hidden state is output; the output layer inputs the second hidden state, and outputs the estimated value of the electric field waveform at any point in the simulation area.

2. The method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator according to claim 1, characterized in that: The excitation source waveform adopts the HEMP double exponential waveform specified in the IEC61000-2-9 standard, and the polarization mode of the excitation signal is horizontal polarization; the surrounding simulated environment is air, and the boundary condition of the simulator is set to an open boundary.

3. The method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator according to claim 1, characterized in that: The three-dimensional coordinates are obtained by one-hot encoding to obtain the three-dimensional coordinate vector; The electric field waveform data is a vector containing the electric field strength value at each moment, which is spliced ​​with the three-dimensional coordinate vector of the measurement space point to obtain the electric field characteristics at the measurement space point; The three-dimensional coordinate vector of the non-measurement space point is concatenated with a zero vector to obtain the electric field characteristics at the non-measurement space point, and the dimension of the zero vector is the same as the dimension of the vector of the electric field waveform data.

4. The method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator according to claim 1, characterized in that: The calculation formula for the first hidden layer is as follows: ； in represents the node hidden state of the vth node, is the set of nodes that have edges with node v, is the weight matrix of the first hidden layer, is the bias term of the first hidden layer, represents the node feature of the u-th node, is a normalizing constant, It is a S-shaped function; The node feature of a node represents the electric field feature at the measurement space point or the electric field feature at the non-measurement space point; ； in represents the first hidden state of the output, represents the node hidden state of the vth node, It means to concatenate the hidden states of all nodes in the graph structure data, N represents the set of all nodes in the graph structure data, is the weight parameter, is the bias parameter, Represents a sigmoid function.

5. The method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator according to claim 4, characterized in that: The calculation formula for the second hidden layer is as follows: ； ； ； ； in, , , represents the weight parameter, , , represents the bias parameter, represents the dot product, where represents the first hidden state output when the tth sequence unit inputs the first hidden layer, and They represent the t-th and t-1-th second hidden states, respectively, n≥t≥1, n represents the total number of sequence units of the sequence data, and when t=1 , tanh is the hyperbolic tangent function, Represents a sigmoid function.

6. The method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator according to claim 5, characterized in that: The calculation formula of the output layer is as follows: ； in represents an estimation matrix of an electric field waveform, a row vector of the electric field waveform estimation matrix represents an electric field waveform vector of a non-measurement point, and the g-th component of the electric field waveform vector represents the electric field strength value at the g-th moment; represents the t-th second hidden state, Indicates the concatenation of the 1st to the nth second hidden states, n represents the total number of sequence units of the sequence data, is the weight parameter, is the bias parameter, Represents a sigmoid function.

7. The method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator according to claim 6, characterized in that: The first hidden layer is pre-trained and connected to the pre-trained output layer during pre-training. The pre-trained output layer inputs the node features of two nodes and outputs the transfer function between the spatial points represented by the two nodes. The calculation formula of the pre-trained output layer is as follows: ； in, It represents the transfer function between nodes i and j. , represents the concatenation of the hidden state of the i-th node and the hidden state of the j-th node, is the weight parameter of the output layer, is the bias parameter of the output layer, represents the S-type function, Represents the transpose of a matrix.

8. The method for simulating a horizontally polarized high-altitude nuclear electromagnetic pulse simulator according to claim 7, characterized in that: The true value calculation formula of the loss function of the first hidden layer pre-training is as follows: ； in, represents the transfer function between node i and node j, represents the angular frequency, and Respectively represent the electric field time domain waveforms corresponding to node i and node j, represents Fourier transform.

9. A horizontally polarized high-altitude nuclear electromagnetic pulse simulator simulation system, characterized in that: It is used to execute a horizontally polarized high-altitude nuclear electromagnetic pulse simulator simulation method as described in any one of claims 1 to 8. A horizontally polarized high-altitude nuclear electromagnetic pulse simulator simulation system comprises: Model building module, used to build 3D electromagnetic pulse simulation models in CST simulation software tools; A data acquisition module, used to collect graph structure data when running full-wave simulation; A data encoding module, used to encode graph structure data into sequence data; The data processing module is used to input the sequence data into the model and output the estimated value of the electric field waveform at any point in the simulation area.

10. A storage medium storing non-transitory computer-readable instructions, characterized in that: When the non-transitory computer-readable instructions are executed by a computer, the steps in the horizontally polarized high-altitude nuclear electromagnetic pulse simulator simulation method as described in any one of claims 1-8 can be executed.

Images