SAR imaging rapid simulation method, device and equipment based on electromagnetic calculation

Abstract

The invention relates to the technical field of radar imaging simulation, in particular to an SAR imaging rapid simulation method, device and equipment based on electromagnetic calculation. According to the method, firstly, an SAR imaging scene is determined, and the scene comprises signal parameters, radar position information and target information; secondly, constructing an LFM signal, and performing fast frequency domain simulation according to an electromagnetic calculation method based on Z-buffer occlusion judgment to obtain RCS data; convolution is carried out on the LFM signal and the RCS data to obtain a radar echo signal, and down-conversion, distance pulse compression, azimuth pulse compression and other operations are carried out on the echo signal in sequence to obtain a two-dimensional SAR image; according to the method, an echo signal of a target is rapidly simulated and SAR imaging is carried out by utilizing a physical optical method based on Z-buffer shielding judgment and an electromagnetic calculation mode combining parallel calculation and difference sweep frequency, the simulation result is more accurate than that of a traditional method based on ideal scattering points, and the speed is higher than that of a traditional electromagnetic algorithm.

Description

Technical Field

[0001] This invention relates to the field of radar imaging technology, and in particular to a rapid simulation method, apparatus, and device for SAR imaging based on electromagnetic calculation. Background Technology

[0002] When electromagnetic waves emitted by a radar system reach a target, they are scattered. The scattering characteristics of electromagnetic waves depend on the target's environment, shape, size, and material. The scattered waves return to the radar receiver in different ways; these returned signals are called radar echoes. Synthetic Aperture Radar (SAR) imaging, based on the electromagnetic scattering and radar echoes, utilizes multiple echo signals to achieve high-resolution imaging of the target. It is widely used in military reconnaissance, environmental monitoring, geological exploration, and disaster assessment. Unlike traditional radar, SAR synthesizes multiple echo data to form a "synthetic aperture," thereby achieving a resolution higher than the actual physical size of the antenna.

[0003] The development of electromagnetic calculation methods can be broadly divided into three key stages: analytical methods, approximation methods, and numerical methods. Analytical methods yield mathematically rigorous results with virtually no error. However, they fail when dealing with complex structures, significantly limiting their application. In such cases, approximation methods can be used. Approximation methods often approximate Maxwell's equations under specific conditions, sacrificing accuracy to obtain a simplified solution. These approximation methods are primarily optical methods, such as physical optics, geometric optics, and uniform geometric diffraction. With the advancement of computing power, full-wave simulation algorithms based on numerical calculations have experienced explosive growth, with the most classic examples being the finite element method, the finite-difference time-domain algorithm, the method of moments, and the fast multipole technique.

[0004] In the development of SAR imaging algorithms for medium and large targets across a wide frequency band, a large amount of echo data is required for testing and verification. Typically, the performance of the radar system is evaluated and optimized by constructing and simulating the echo signals of ideal scattering points. However, the signals are not realistic, resulting in inaccurate SAR images. A more precise approach is to perform 3D modeling of the target, then use electromagnetic calculation methods to simulate the echo, and finally perform SAR imaging simulation. Common electromagnetic calculation methods, such as the method of moments (MoM), while highly accurate, have high computational complexity and low efficiency; fast multi-machine methods offer high accuracy, require less memory, and are fast, but need to consider iterative convergence issues; domain decomposition methods are fast, but their application scenarios are limited; and the bouncing ray method offers high computational accuracy and is applicable to various situations, but its computational process is complex and consumes significant computational resources. Therefore, new technologies are needed to improve the speed and accuracy of radar cross-section (RCS) simulation.

[0005] NVIDIA in the US has developed a GPU ray tracing method based on the bouncing ray method. Its advantages include high computational accuracy and fast GPU computing speed, but the CUDA core library called in the code cannot be modified. Poland has developed a method between physical optics and ideal scattering points, simulating echo signals by discretizing surfaces into ideal points, approximating reflection coefficients, and performing simple occlusion checks. Its advantages include high computational efficiency and ease of use, but it is not a true electromagnetic simulation; its accuracy is only slightly higher than the ideal scattering point model. Germany developed POV-Ray, a ray tracing engine, in 2016. It is currently the only free SAR imaging simulation software based on electromagnetic simulation, and it is also relatively fast.

[0006] In the prior art, the invention with patent publication number CN 109188384 A, entitled "Electromagnetic Simulation Method for Dynamic Observation of Space Target Echoes," discloses an electromagnetic simulation method for dynamic observation of space target echoes. This method calculates the dynamic observation angle and optimal observation time of the space target based on the positions of the space target and the ground observation station. At the dynamic observation angle during the optimal observation time, the electromagnetic scattering of the space target is calculated using the large-area physical optics method to obtain the space target echo and form an image. This achieves effective electromagnetic simulation and imaging of the space target under different dynamic observation angles, reducing computational load and enabling multiple observation angles. However, it still suffers from the problem that time costs need to be reduced.

[0007] In the prior art, patent publication number CN 118152928 A, entitled "A Method for Forward Modeling of Multiple Scattering Centers of Radar Targets Based on a Mesh Model," discloses a method for forward modeling of multiple scattering centers of radar targets based on a mesh model. This method utilizes a parametric model of the target's scattering centers, along with bouncing ray technology and forward modeling techniques based on multiple scattering centers, to perform forward modeling of radar targets. After modeling, electromagnetic calculations can be performed using the parametric representation of the target's scattering centers, enabling electromagnetic calculations for large targets or scenes. This method simplifies the complexity of electromagnetic calculations and improves simulation efficiency, but it suffers from insufficient accuracy. Summary of the Invention

[0008] In order to overcome the shortcomings of the prior art, the present invention aims to propose a rapid simulation method, device and equipment for SAR imaging based on electromagnetic calculation. This method improves the accuracy and speed of echo signal simulation by combining physical optics based on Z-buffer occlusion judgment with electromagnetic calculation method that combines parallel computing and difference frequency sweeping.

[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows: Firstly, a rapid simulation method for SAR imaging based on electromagnetic calculations includes the following steps: S1: Target geometric modeling: Determine the scene information of SAR imaging and establish a target geometric model. The scene information includes signal parameters, radar position information, and target information. S2: Frequency domain electromagnetic simulation: The target geometric model described in step S1 is simulated in the frequency domain using the physical optics method to obtain the frequency domain data H(jω) of the RCS of the target geometric model. S3: Generate echo signal: Transform the frequency domain data H(jω) of the RCS obtained in step S2 to the time domain to obtain the time domain data of the RCS. (t), the linear frequency modulated (LFM) signal calculated using the signal parameters described in step S1. and the time domain data of the RCS (t) is convolved to obtain the echo signal. Among them, the linear frequency modulated signal is a transmitted signal; S4: SAR imaging: Analyzing the echo signal described in step S3 The two-dimensional SAR image is obtained by sequentially performing down-conversion, range pulse compression, and azimuth pulse compression. When a plane wave or a point source excites an ideal conductor (PEC) scatterer, the incident field is assumed to be ( , The specific process of the physical optics method described in step S2 is as follows: S2.1: Reconstruct the surface of the target geometric model described in step S1 using M (M=1, 2, ..., m) triangular units; S2.2: Inspired by the depth buffer Z-buffer technique in computer graphics, this invention proposes a Z-buffer algorithm for occlusion detection by recording the depth of each visible pixel. The occlusion judgment of the M (M=1, 2, ..., m) triangular units in step S2.1 is performed by the deep buffer Z-buffer algorithm to obtain the set of bright area triangles. This process realizes the application of electromagnetic calculation to SAR imaging and improves accuracy. S2.3: The induced current generated on the surface of the bright triangular set described in step S2.2 is calculated using RWG basis functions, thereby discretizing the current distribution. Then, the scattered field is obtained by integrating the induced current and its divergence using Green's function in free space. Further calculation of the scattered field The frequency domain data of RCS can then be obtained. induced current Scattered field Frequency domain data of RCS The calculation formula is as follows: In the formula, For induced current, It is a scattered field (induced current radiation field). It is the normal vector. For the incident magnetic field, Angular frequency, Permeability, Where is the dielectric constant. This represents the vector from the field point to the source point. The distance from the field point to the source point. This is the Green's function.

[0010] Furthermore, the occlusion determination in step S2.2 specifically includes the following sub-steps: S2.2.1: Divide the background plane in the scene information described in step S1 into... Each grid is used with 1 grid. It indicates that 1 of them ≤ 1≤ ≤ Then, the M (M=1, 2, ..., m) triangular units described in step S1 are projected onto the background plane from the light source position, and an array is defined. Represents blocking different grids The triangular unit number; S2.2.2: For the first triangular unit of the projection described in step S2.2.1, if the mesh on the background plane... If the grid is occluded by the first triangular unit, then assign a value to that grid. corresponding array Assign a value of 1, and record the number of projected meshes of the first triangular unit on the background plane. For the Mth (M=2, 3, ..., m) triangular element of the projection described in step S2.2.1, for the occluded and unassigned mesh... Directly give the grid corresponding array The value is assigned to M, where M = 2, 3, ..., m, for occluded but already assigned meshes. Compare the M-th (M=2, 3…m) triangular unit with the... ( The array represents the distances from each of the <M> triangular units to the light source. The value is assigned to the triangular element number closer to the light source, and then the number of projected grids of that triangular element on the background plane is recorded. Repeat this process until the projection operation is completed for all M (M=2, 3, ..., m) triangular elements; S2.2.3: Calculate the different assignment arrays described in step S2.2.2. Corresponding grid quantity Where M = 1, 2, 3, ..., m, if If the triangle is bright, then the triangle is a bright triangle, and all the bright triangles form a set of bright triangles, while the remaining triangles are occluded.

[0011] Furthermore, the calculation process of steps S2.2 and S2.3 is completed using MPI parallel computing: Assuming that P occlusion judgment processes are required in step S2.2, when using Q processes in parallel, all occlusion judgment processes are divided into Q groups, that is, each group needs to perform P / Q occlusion judgments. Then, the main process accumulates the projection matrices of each process, and finally quickly judges the occlusion result through the projection matrix. In step S2.3, the bright area triangles are divided into Q groups, with approximately equal numbers of triangles in each group. Each process independently calculates a set of integrals, and finally, the main process accumulates all the integral results to obtain the final electric field. The key to parallel computing is to distribute the overall task evenly among the processes. We mainly perform MPI parallel processing on the time-consuming steps S2.2 and S2.3 of the aforementioned physical optics method, namely occlusion detection and far-field calculation. Theoretically, MPI parallelism with tens of thousands of cores can be achieved; in practice, parallelism with 128 cores is sufficient for fast simulation, typically within 12 hours.

[0012] Furthermore, the frequency domain data of the RCS described in step S2.3 is approximated using a rational interpolation polynomial. Then, the zeros and poles of the rational interpolation polynomial are analyzed, and the causality of the rational interpolation polynomial is corrected through analytical methods. After using difference scanning, accurate RCS data across the entire frequency band can be obtained quickly and stably using data from only a few frequency points within the band.

[0013] Furthermore, the linear frequency modulation signal mentioned in step S3 The expression is as follows: in, , For the center frequency, For all-time variables, For pulse width, To adjust the frequency; The echo signal The expression is as follows: Furthermore, in step S4, the selected reference signal during the down-conversion... The expression is as follows: After the echo signal is demodulated The expression is as follows: in, In order to receive signals, This refers to the attenuation of the signal after it travels through space. The distance from the target to the launch point. It is the speed of light.

[0014] Furthermore, in the distance pulse compression described in step S4, the down-converted echo signal is processed using matched filtering technology. The processing involves using convolution operations to perform matched filtering, which converts long signal pulses into signal streams. Compressed into a short-time pulse ; Assuming the radar travels from point A to point B at a constant speed in a straight line, transmitting a linear frequency modulated signal with a pulse interval of period T, the received echo signal, after demodulation processing, can be represented in a mathematical form containing target information: in, It is a directional time variable. It is a distance-to-time variable. It is the distance from each sampling point to the target point; The system function of the selected matched filter Represented as: in, It is the reference slope distance, which is the minimum distance between all sampling points and the observation range; Represented as: Furthermore, step S4, azimuth pulse compression, sequentially includes three sub-steps: scene mesh partitioning, migration curve determination, and signal coherent accumulation. Specifically, scene mesh partitioning involves dividing the detection area into... Each grid point represents a pixel in the imaging area, and the size and distribution of the grid points depend on the required resolution and the size of the imaging area. The migration curve is determined by calculating the distance change of each grid point at different azimuth times based on the relationship between the radar platform's motion trajectory and the target's position. Coherent signal accumulation specifically involves accumulating the echo signals over time to improve the signal-to-noise ratio and resolution. Specifically, for each grid point, echo signals at different azimuth times are coherently accumulated along its corresponding distance migration curve. Any grid point in the scene mesh partitioning Corresponding echo signal Represented as: Where Mslow is the pulse number, i.e., the number of sampling points on the aperture. For grid points The distance to the nth sampling point.

[0015] Secondly, a rapid simulation device for SAR imaging based on electromagnetic calculation includes: Target geometry modeling module: determines the scene information of SAR imaging and establishes the target geometry model, wherein the scene information includes signal parameters, radar position information, and target information; Frequency domain electromagnetic simulation module: The target geometric model is simulated in the frequency domain using the physical optics method to obtain the frequency domain data H(jω) of the RCS of the target geometric model; Echo signal generation module: Transforms the frequency domain data H(jω) of the RCS to the time domain to obtain the time domain data of the RCS. (t), the linear frequency modulated (LFM) signal calculated using the aforementioned signal parameters. and the time domain data of the RCS (t) is convolved to obtain the echo signal; SAR imaging module: The echo signal is sequentially down-converted, range pulse compressed, and azimuth pulse compressed to obtain a two-dimensional SAR image.

[0016] Thirdly, an electronic device includes a memory and a processor, wherein: Memory: Used to store the computer program that implements the electromagnetic calculation-based SAR imaging rapid simulation method; the memory includes: USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk or optical disk and other media that can store program code; Processor: Used to implement the electromagnetic calculation-based SAR imaging fast simulation method when executing the computer program.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. In the method of the present invention, step S2 uses an electromagnetic calculation method to replace the traditional method based on ideal scattering points, thereby improving the accuracy of the echo signal of the simulated target and facilitating subsequent high-precision imaging of the target area.

[0018] 2. In the method of the present invention, through steps S2 and S3, the target echo signal is quickly simulated by using a physical optics method based on Z-buffer occlusion judgment, parallel computing, and electromagnetic calculation method combining difference frequency sweeping. This ensures accuracy while accelerating the speed of electromagnetic calculation.

[0019] 3. In the method of this invention, rapid simulation of SAR imaging based on electromagnetic calculation is achieved through a combination of steps S2, S3, and S4. This method allows for the analysis of imaging effects under different target, material, and environmental conditions, thereby enabling in-depth research into the relationship between scattering characteristics and imaging results. Therefore, it has significant advantages in simulating and analyzing complex scenes. For example, in complex scattering scenes such as the ground surface, buildings, or vegetation, electromagnetic calculation can accurately capture complex phenomena such as multipath scattering, coherent superposition, and polarization effects, providing reliable data support for subsequent high-resolution imaging and feature extraction.

[0020] In summary, the method of this invention utilizes a physical optics method based on Z-buffer occlusion judgment and an electromagnetic calculation method combining parallel computing and differential frequency sweeping to quickly simulate the target's echo signal and perform SAR imaging. By reconstructing the radar image of the target area through a back projection imaging algorithm, including signal processing methods such as range compression and azimuth pulse compression, the results are more accurate than traditional methods based on ideal scattering points and faster than traditional electromagnetic algorithms. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the method.

[0022] Figure 2(a) is a schematic diagram of the triangular unit and the projected background plane in the Z-buffer algorithm of this method.

[0023] Figure 2(b) is a schematic diagram of the overlapping projections of two triangular units in the Z-buffer algorithm of this method.

[0024] Figure 3(a) is a schematic diagram of the scene configuration of Embodiment 1 of the present invention.

[0025] Figure 3(b) is a schematic diagram of the time domain curve of the transmitted LFM signal in Embodiment 1 of the present invention.

[0026] Figure 3(c) is a schematic diagram of the spectrum amplitude corresponding to the transmitted LFM signal in Embodiment 1 of the present invention.

[0027] Figure 3(d) is a schematic diagram of the echo signal obtained by simulation in Embodiment 1 of the present invention.

[0028] Figure 4(a) is a schematic diagram of the stealth drone structure of Embodiment 2 of the present invention.

[0029] Figure 4(b) shows the imaging result of the stealth drone in Embodiment 2 of the present invention using this method.

[0030] Figure 5 This is the imaging configuration of Embodiment 2 of the present invention.

[0031] Figure 6(a) is a schematic diagram of the conventional 241 ideal scattering points of the stealth drone in Embodiment 2 of the present invention.

[0032] Figure 6(b) shows the imaging results of the stealth drone of Embodiment 2 of the present invention under the traditional 241 ideal scattering point model.

[0033] Figure 7(a) is a schematic diagram of the conventional 8845 ideal scattering points of the stealth drone in Embodiment 2 of the present invention.

[0034] Figure 7(b) shows the imaging results of the stealth drone of Embodiment 2 of the present invention under the traditional 8845 ideal scattering point model. Detailed Implementation

[0035] The present invention will now be described in further detail with reference to the accompanying drawings: Firstly, such as Figure 1 As shown, this invention provides a rapid simulation method for radar SAR imaging based on electromagnetic computation, comprising four parts: geometric modeling, frequency domain electromagnetic simulation, echo signal generation, and SAR imaging. First, the SAR imaging scene is determined, and a geometric model is established. Second, frequency domain electromagnetic simulation is performed based on the geometric model, specifically using a combination of parallel physical optics and differential scanning to obtain frequency domain data of the RCS (radar cross-section). Next, the frequency domain data is subjected to FFT to obtain the time domain response. The echo signal is obtained by convolving the transmitted signal (i.e., the linear frequency modulated signal) with the echo signal (t); finally, the echo signal is processed to perform SAR imaging.

[0036] The SAR imaging rapid simulation method of this invention specifically includes the following steps: S1: Target geometric modeling: Determine the scene information of SAR imaging and establish a target geometric model. The scene information includes signal parameters, radar position information, and target information. S2: Frequency domain electromagnetic simulation: The target geometric model described in step S1 is simulated in the frequency domain using the physical optics method to obtain the frequency domain data H(jω) of the RCS of the target geometric model. S3: Generate echo signal: Transform the frequency domain data H(jω) of the RCS obtained in step S2 to the time domain to obtain the time domain data of the RCS. (t), the linear frequency modulated (LFM) signal calculated using the signal parameters described in step S1. and the time domain data of the RCS (t) is convolved to obtain the echo signal. Among them, the linear frequency modulated signal is a transmitted signal; S4: SAR imaging: Analyzing the echo signal described in step S3 The two-dimensional SAR image is obtained by sequentially performing down-conversion, range pulse compression, and azimuth pulse compression. Under high-frequency conditions, when using the physical optics method to calculate the induced current, the following three basic premises must be met: the target size is significantly smaller than the distance between the observation point and the target, and the target size is significantly larger than the wavelength; induced current exists only in the region directly hit by the incident wave (bright region); and the characteristics of the induced current in the bright region are the same as the current characteristics on an infinitely large plane tangent to the incident point and the surface. When a plane wave or a point source excites an ideal conductor (PEC) scatterer, the incident field is assumed to be ( , The specific process of the physical optics method described in step S2 is as follows: S2.1: Reconstruct the surface of the target geometric model described in step S1 using M (M=1, 2, ..., m) triangular units; S2.2: Inspired by the depth buffer Z-buffer technique in computer graphics, this invention proposes a Z-buffer algorithm for occlusion detection by recording the depth of each visible pixel. The occlusion judgment of the M (M=1, 2, ..., m) triangular units in step S2.1 is performed by the deep buffer Z-buffer algorithm to obtain the set of bright area triangles. This process realizes the application of electromagnetic calculation to SAR imaging and improves accuracy. S2.3: The induced current generated on the surface of the bright triangular set described in step S2.2 is calculated using RWG basis functions, thereby discretizing the current distribution. Then, the scattered field is obtained by integrating the induced current and its divergence using Green's function in free space. Further calculation of the scattered field The frequency domain data of RCS can then be obtained. induced current Scattered field Frequency domain data of RCS The calculation formula is as follows: In the formula, For induced current, It is a scattered field (induced current radiation field). It is the normal vector. For the incident magnetic field, Angular frequency, Permeability, Where is the dielectric constant. This represents the vector from the field point to the source point. The distance from the field point to the source point. This is the Green's function.

[0037] Furthermore, the occlusion determination in step S2.2 specifically includes the following sub-steps: S2.2.1: As shown in Figure 2(a), the background plane in the scene information described in step S1 is divided into... Each grid is used with 1 grid. It indicates that 1 of them ≤ 1≤ ≤ Then, the M (M=1, 2, ..., m) triangular units described in step S1 are projected onto the background plane from the light source position, and an array is defined. Represents blocking different grids The triangular unit number; S2.2.2: As shown in Figure 2(b), for the first triangular unit of the projection described in step S2.2.1, if the grid on the background plane... If the grid is occluded by the first triangular unit, then assign a value to that grid. corresponding array Assign a value of 1, and record the number of projected meshes of the first triangular unit on the background plane. For the Mth (M=2, 3, ..., m) triangular element of the projection described in step S2.2.1, for the occluded and unassigned mesh... Directly give the grid corresponding array The value is assigned to M, where M = 2, 3, ..., m, for occluded but already assigned meshes. Compare the M-th (M=2, 3…m) triangular unit with the... ( The array represents the distances from each of the <M> triangular units to the light source. The value is assigned to the triangular element number closer to the light source, and then the number of projected grids of that triangular element on the background plane is recorded. Repeat this process until the projection operation is completed for all M (M=2, 3, ..., m) triangular elements; S2.2.3: Calculate the different assignment arrays described in step S2.2.2. Corresponding grid quantity Where M = 1, 2, 3, ..., m, if If the triangle is bright, then the triangle is a bright triangle, and all the bright triangles form a set of bright triangles, while the remaining triangles are occluded.

[0038] Example 1 The echo simulation of this invention is verified using Matlab software. As shown in Figure 3(a), assume there is an antenna and a metal triangular target in free space. The vertex coordinates of the triangle are (100.9960, 100.9935, 0.0061), (100.9956, 100.9956, 0.0083), and (100.9935, 100.9960, 0.0061). The antenna's position coordinates are (0, 0, 100). All coordinates are in meters (m). At a specific moment, the antenna transmits a plane wave with a carrier frequency of 65 GHz and a bandwidth of 500 MHz. After being scattered by the target, the antenna receives the echo. The time-domain transmitted signal and its spectrum are shown in Figures 3(b) and 3(c). The echo signal calculated using the electromagnetic algorithm of this invention when the target is a triangle is shown in Figure 3(d). The closest distance between the antenna and the target is 174.3593m, so the estimated time delay is 1.1632 microseconds. Comparing the pulse start points in Figure 3(b) and Figure 3(d), the latter's time delay is 1.1631 microseconds, which matches the estimated time delay well, indicating that the echo simulation method of the present invention is correct.

[0039] Example 2 The algorithm of this invention is used below to image the stealth UAV shown in Figure 4(a) using Matlab software. The configuration during imaging is as follows. Figure 5 As shown, the radar's initial observation position is (0,0,100), and the endpoint position is (0,7.4019,100), with a total of 128 observation points. The radar's flight speed is 2000 m / s, the center point coordinates of the imaging area are (100,120,0), and the area size is 65m × 78m. The electromagnetic wave carrier frequency is 10 GHz, and the bandwidth is 500 MHz. Therefore, the azimuth resolution is 0.0472m, and the range resolution is 0.2998m. Figure 4(b) shows the imaging results of the echo signal simulation method calculated based on physical optics, where the outline of the UAV can be clearly seen, indicating that the algorithm is very accurate.

[0040] In traditional methods for simulating echo signals from ideal scattering points, determining the optimal number of points to approximate the contour of a large target remains undetermined. As shown in Figures 6(a) and 6(b), the outer surface of a stealth UAV is discretized into sets of 241 and 8845 point targets, respectively, with imaging results shown in Figures 7(a) and 7(b). The more points used, the more blurred the UAV's outer contour becomes in the image, indicating a limitation of this simulation method.

[0041] In summary, the echo simulation method based on electromagnetic calculations is more accurate than the traditional method for simulating echo signals from ideal scattering points.

[0042] The working principle of this invention is as follows: In this method, the scene for SAR imaging is first determined, including signal parameters, radar position information, and target information. Then, the transmitted LFM signal is constructed, and rapid frequency domain simulation is performed according to the electromagnetic calculation method of this invention to obtain RCS data. Next, the transmitted signal and RCS data are convolved to obtain the radar echo signal. The echo signal is then subjected to down-conversion, range pulse compression, and azimuth pulse compression operations in sequence. The azimuth pulse compression can be further subdivided into three parts: scene grid subdivision, migration curve determination, and signal coherent accumulation. Finally, a two-dimensional SAR image is obtained.

Claims

1. A rapid simulation method for SAR imaging based on electromagnetic calculation, characterized in that, Includes the following steps: S1: Determine the scene information for SAR imaging and establish a target geometric model. The scene information includes signal parameters, radar location information, and target information. S2: Perform frequency domain electromagnetic simulation on the target geometric model described in step S1 using the physical optics method to obtain the frequency domain data H(jω) of the RCS of the target geometric model; S3: Transform the frequency domain data H(jω) of the RCS obtained in step S2 to the time domain to obtain the time domain data of the RCS. (t) will be the linear frequency modulated signal calculated using the signal parameters described in step S1. and the time domain data of the RCS (t) is convolved to obtain the echo signal. ; S4: The echo signal described in step S3 The two-dimensional SAR image is obtained by sequentially performing down-conversion, range pulse compression, and azimuth pulse compression. The specific process of the physical optics method described in step S2 is as follows: S2.1: Reconstruct the surface of the target geometric model described in step S1 using M (M=1, 2, ..., m) triangular units; S2.2: The occlusion judgment of the M (M=1, 2, ..., m) triangular units mentioned in step S2.1 is performed by the depth buffer Z-buffer algorithm to obtain the set of bright area triangles; S2.3: Calculate the induced current generated on the surface of the bright region triangle set described in step S2.2 using RWG basis functions, and then obtain the scattered field by integrating the induced current and its divergence using Green's functions in free space. The frequency domain data of RCS can then be calculated. induced current Scattered field Frequency domain data of RCS The calculation formula is as follows: In the formula, For induced current, For the scattered field, It is the normal vector. For the incident magnetic field, Angular frequency, Permeability, Where is the dielectric constant. This represents the vector from the field point to the source point. The distance from the field point to the source point. This is the Green's function.

2. The SAR imaging rapid simulation method as described in claim 1, characterized in that, Step S2.2, the occlusion determination, specifically includes the following sub-steps: S2.2.1: Divide the background plane in the scene information described in step S1 into... Each grid is used with 1 grid. It indicates that 1 of them ≤ 1≤ ≤ Then, the M (M=1, 2, ..., m) triangular units described in step S1 are projected onto the background plane from the light source position, and an array is defined. Represents blocking different grids The triangular unit number; S2.2.2: For the first triangular unit of the projection described in step S2.2.1, if the mesh on the background plane... If the grid is occluded by the first triangular unit, then assign a value to that grid. corresponding array Assign a value of 1, and record the number of projected meshes of the first triangular unit on the background plane. ; For the Mth (M=2, 3, ..., m) triangular element of the projection described in step S2.2.1, for the occluded and unassigned mesh... Directly give the grid corresponding array The value is assigned to M, where M = 2, 3, ..., m, for occluded but already assigned meshes. Compare the M-th (M=2, 3…m) triangular unit with the... ( The array represents the distances from each of the <M> triangular units to the light source. The value is assigned to the triangular element number closer to the light source, and then the number of projected grids of that triangular element on the background plane is recorded. Repeat this process until the projection operation is completed for all M (M=2, 3, ..., m) triangular elements; S2.2.3: Calculate the different assignment arrays described in step S2.2.

2. Corresponding grid quantity Where M = 1, 2, 3, ..., m, if If the triangular unit is a bright area triangle, then all the bright area triangles form a set of bright area triangles.

3. The SAR imaging rapid simulation method as described in claim 1, characterized in that, The calculation process of steps S2.2 and S2.3 is completed using MPI parallel computing.

4. The SAR imaging rapid simulation method as described in claim 1, characterized in that, The frequency domain data of the RCS described in step S2.3 are approximated using rational interpolation polynomials. Then, the zeros and poles of the rational interpolation polynomial are analyzed, and the causality of the rational interpolation polynomial is corrected through analytical methods.

5. The SAR imaging rapid simulation method as described in claim 1, characterized in that, The linear frequency modulation signal in step S3 The expression is as follows: in, , For the center frequency, For all-time variables, For pulse width, To adjust the frequency; The echo signal The expression is as follows: 。 6. The SAR imaging rapid simulation method as described in claim 1, characterized in that, In step S4, the selected reference signal during the down-conversion process The expression is as follows: After the echo signal is demodulated The expression is as follows: in, In order to receive signals, This refers to the attenuation of the signal after it travels through space. The distance from the target to the launch point. It is the speed of light.

7. The SAR imaging rapid simulation method as described in claim 1, characterized in that, In step S4, during distance pulse compression, matched filtering is used to process the down-converted echo signal. The processing involves using convolution operations to perform matched filtering, which converts long signal pulses into signal streams. Compressed into a short-time pulse ; Assuming the radar travels from point A to point B at a constant speed in a straight line, transmitting a linear frequency modulated signal with a pulse interval of period T, the received echo signal, after demodulation processing, can be represented in a mathematical form containing target information: in, It is a directional time variable. It is a distance-to-time variable. It is the distance from each sampling point to the target point; The system function of the selected matched filter Represented as: in, It is the reference slope distance, which is the minimum distance between all sampling points and the observation range; Represented as: 。 8. The SAR imaging rapid simulation method as described in claim 1, characterized in that, Step S4, the azimuth pulse compression, includes three sub-steps: scene mesh partitioning, migration curve determination, and signal coherent accumulation. In the scene mesh partitioning, any mesh point... Corresponding echo signal Represented as: Where Mslow is the pulse number, i.e., the number of sampling points on the aperture. For grid points The distance to the nth sampling point.

9. An apparatus for implementing the rapid simulation method for SAR imaging as described in any one of claims 1 to 8, characterized in that, include: Target geometry modeling module: determines the scene information of SAR imaging and establishes the target geometry model, wherein the scene information includes signal parameters, radar position information, and target information; Frequency domain electromagnetic simulation module: The target geometric model is simulated in the frequency domain using the physical optics method to obtain the frequency domain data H(jω) of the RCS of the target geometric model; Echo signal generation module: Transforms the frequency domain data H(jω) of the RCS to the time domain to obtain the time domain data of the RCS. (t), the linear frequency modulated signal calculated using the signal parameters. and the time domain data of the RCS (t) is convolved to obtain the echo signal; SAR imaging module: The echo signal is sequentially down-converted, range pulse compressed, and azimuth pulse compressed to obtain a two-dimensional SAR image.

10. An electronic device comprising a memory and a processor, wherein: Memory: for storing computer programs that implement the SAR imaging rapid simulation method as described in any one of claims 1 to 8; Processor: Used to implement the SAR imaging fast simulation method as described in any one of claims 1 to 8 when executing the computer program.

Images