A fast semi-analytical calculation method for scattering of composite targets based on a non-uniform mesh model

Abstract

This invention discloses a fast calculation method for composite target scattering based on non-uniform mesh generation and the PO-EEC-IEM-ASC-FPM method: large mesh generation is performed on the target and rough surface, and PO-EEC and IEM-ASC are used for solving; for the target mesh with coupling effect, the mesh is refined and FPM is used for solving, which minimizes the unknowns of the model; at the same time, when performing coupling solution, only the coupling scattering amplitude of one refined coupling surface element is calculated, and all coupling amplitudes are obtained analytically, which reduces the number of repeated calculations and significantly improves the calculation efficiency of the total scattering field of composite scene.

Description

Technical Field

[0001] This invention relates to the field of electromagnetic scattering calculation technology, specifically to an efficient calculation method for the composite scattering characteristics of a target in a highly complex, electrically charged environment. Background Technology

[0002] The rapid and accurate calculation of target composite scattering in complex environments is a crucial problem urgently needing to be solved in the field of electromagnetic simulation. In practical applications, complex radar targets coupled with rough terrain backgrounds are one of the most common composite scenarios. However, the complex structure of the target and its background, with electrical dimensions much larger than the incident wavelength (when the electrical size reaches thousands of wavelengths, it is called ultra-electrically large), and the terrain background exhibiting multi-scale, random, and rough characteristics, pose a significant challenge to the accurate and efficient electromagnetic simulation of the scattering characteristics of composite scenarios, especially the coupled scattering between the target and the rough background.

[0003] Current techniques for simulating the combined scattering of a rough background and a target include four main categories: full-wave method, high-frequency algorithm, high-frequency-full-wave hybrid algorithm, and high-frequency-high-frequency hybrid algorithm. Typical techniques for each category are as follows:

[0004] 1. Full-wave numerical algorithms include the method of moments (MoM), finite-difference time-domain (FTDT), and finite element method (FEM). In order to improve the computational efficiency of traditional full-wave methods, techniques such as multilevel fast multipole (MLFMA), domain decomposition (DDM), hybrid finite element-boundary element method, and multilevel fast multipole combined element technique (FE-BI-MLFMA) have been proposed.

[0005] 2. High-frequency methods include the bouncing ray method (SBR), geometric optics method (GO), physical optics method (PO), integral equation method (IEM), Kirchhoff approximation method (KA), perturbation method (SPM), four-path model (FPM), and iterative physical optics method (IPO).

[0006] 3. High-frequency-full-wave hybrid algorithms are hybrid algorithms proposed to improve the efficiency of full-wave algorithms. Typical high-frequency-full-wave hybrid algorithms include the KA-MoM method which combines Kirchhoff's approximation method and the method of moments, the PO-MLFMA method which combines physical optics method and multilayer fast multipole method, and the FE-BI-KA method which combines finite element method, boundary element method and Kirchhoff method.

[0007] 4. High-Frequency Hybrid Methods: Since high-frequency methods typically have a limited range of applicability, these methods can broaden the applicability of a single high-frequency method. Typical examples include the PO-SBR method, which combines physical optics and bouncing ray methods; the GO, PO, and PTD-CWMFSM methods, which combine geometric optics, physical optics, physical diffraction theory, and surface-based and multipath models with capillary wave correction; and the PO-IPO method, which combines physical optics and iterative physical optics.

[0008] Although the full-wave numerical method can meet the accuracy requirements for simulating the scattering characteristics of composite targets, it requires a huge amount of computing resources when simulating ultra-large electric composite scene models, and the calculation process is time-consuming. For example, using the method of moments (SDIE-MoM) of self-dual integral equations to calculate a 100λ sea surface-ship composite scene, the calculation time for each composite target model is 37 minutes. Therefore, the full-wave method is difficult to achieve efficient simulation of the scattering characteristics of ultra-large electric composite targets and cannot meet the needs of actual engineering.

[0009] Compared to full-wave numerical methods, high-frequency methods have an advantage in efficiency for solving composite scattering simulations. However, high-frequency methods have certain limitations in applicability. For example, the PO method and IEM method can only calculate single scattering between the target and the rough surface, while multiple scatterings between the target and the rough surface require calculation using high-frequency methods such as SBR and IPO methods. There is a lack of high-frequency algorithms that can directly calculate composite targets.

[0010] The full-wave-high-frequency hybrid algorithm uses the full-wave algorithm to calculate the target and coupled scattering, and the high-frequency method to calculate the scattering of the rough surface. While ensuring accuracy, it improves the computational efficiency to a certain extent. For example, when using the KA-MoM method to calculate the target of the rough surface and the missile above it, the efficiency is 6.6 times higher than that of the full-wave method.

[0011] The high-frequency-high-frequency hybrid algorithm utilizes different high-frequency methods to address different echo components, achieving a further improvement in efficiency compared to the full-wave-high-frequency hybrid algorithm. Taking the PO-SBR hybrid method as an example, its efficiency in calculating the bistatic scattering characteristics of a composite scene of sea surface and ship with an electrical size of 200 wavelengths is 91.8 times higher than the method of moments. Therefore, the high-frequency-high-frequency hybrid algorithm is more suitable for the rapid calculation of scattering from ultra-large electrical composite targets.

[0012] Typically, high-frequency-high-frequency hybrid algorithms divide the echoes of complex scenes into three parts: 1. Independent target echoes; 2. Independent scene echoes; 3. Multiple scattering components between the target and the scene. The essence of the hybrid algorithm is to use targeted scattering models for simulation and correction of different echo components. For example, in the PO-SBR method, the single scattering component of the target is calculated using the PO method; the single scattering component of the rough background is calculated using a surface model with PO correction amplitude; and the coupled scattering component between the target and the rough background is calculated using SBR. In the GO\PO\PTD-CWMFSM hybrid algorithm, GO\PO\PTD is used to calculate the independent target scattering component, CWMFSM is used to calculate the background rough surface scattering component, and a four-path model is used to calculate the coupled scattering component.

[0013] However, when using ray-based methods such as SBR and FPM to solve for coupled scattering between the target and the background, it is necessary to ensure that the coupled mesh of the target has an electrically small size (typically, the mesh size or scattering center length that produces the coupling effect needs to be less than 1 / 5 of the wavelength) to reduce the coupled scattering path error caused by mesh / scattering center height errors, which increases the number of meshes and the number of unknowns. When using current-based iterative methods such as IPO to solve for coupled scattering between the target and the background, the influence of the current of each surface element must be continuously considered, resulting in higher computational complexity and longer computation time, which seriously reduces the simulation efficiency of electromagnetic scattering of ultra-large composite targets. Therefore, efficient and accurate calculation of electromagnetic scattering in ultra-large composite scenarios remains an urgent scientific problem to be solved. Summary of the Invention

[0014] In view of this, based on the electromagnetic scattering characteristics of composite scenes, this invention proposes a non-uniform-semi-analytical composite scene mesh model. It uses hybrid physical optics and equivalent current method (PO-EEC), surface element scattering center model (IEM-ASC) based on integral equation method and four-path model (FPM) to simulate independent target scattering, coarse background scattering and coupled scattering components in composite scenes. It can be applied to rapid simulation of radar echoes of complex targets in ground environment and radar remote sensing technology.

[0015] This disclosure provides a fast semi-analytical calculation method for scattering of composite targets based on a non-uniform mesh model, including the following steps:

[0016] Acquire or construct an independent target mesh model and a rough background terrain mesh model;

[0017] Determine the edge structure of the target mesh;

[0018] The scattered echo of an independent target is established using a hybrid physical optics method and an equivalent current method.

[0019] Echoes of rough background terrain are established using a surface element model based on the integral equation method.

[0020] The mesh in the coupling region of the target mesh is uniformly refined, and a semi-analytical four-path model is used to calculate the multiple scattering between the target and the background;

[0021] The scattered echoes of the independent target, the echoes of the rough background terrain, and the multiple scatterings between the target and the background are summed to obtain the scattered field of the composite target.

[0022] Furthermore, the step of determining the target mesh edge structure specifically includes the following methods:

[0023] Read the target mesh file and determine the vertex coordinates corresponding to all meshes;

[0024] Solve for each edge vector of each grid cell, and use the edge vectors to solve for the out-of-grid normal vector;

[0025] Iterate through all grids that have two identical vertex numbers and define them as adjacent grids;

[0026] Solve for the angle between the outward normal vectors of each pair of adjacent meshes. If the angle satisfies certain conditions, it is determined to be an edge structure. Obtain the outward normal vectors of the edge endpoints and the two lateral face elements of the mesh.

[0027] Furthermore, the step of establishing the scattered echo of an independent target using the hybrid physical optics method and the equivalent current method specifically includes:

[0028] The independent target scattered echoes, divided into a mesh model, are equivalent to the superposition of all mesh surface scattering components and all edge structure diffraction components:

[0029]

[0030] in, For the scattered echo of an independent target, Let i be the surface scattering field of the i-th grid. Let I be the diffraction field of the j-th edge, where I and J represent the number of grid cells and the number of edges, respectively.

[0031] The surface scattering components are calculated using the simplified PO contour integral formula:

[0032]

[0033] in, Represents the outside normal vector of the mesh; Indicates the receiving direction of the unit; T is The projected length on the plane of the flat plate, where M is the total number of edges of the flat plate. It is the position vector of the source point. This represents the length and direction of the m-th side of the plate. For tablet and Vertical unit direction;

[0034] Calculate the diffraction components of the edge structure using EEC:

[0035]

[0036]

[0037] in, Let η0 be the direction vector at the edge, and η0 be the free space impedance. and These represent the vertical and horizontal polarization directions, respectively. and These are the start and end points of the edge, respectively. Let I be the midpoint of the edge. PTD and M PTD These are the diffraction current and diffraction magnetic current at the edge, respectively.

[0038] Furthermore, the method for establishing the echo of the rough background terrain using a surface element model based on the integral equation method includes:

[0039] The rough surface is divided into electrically large grids with the imaging resolution as the size. Each grid can be regarded as a point scattering center, and the amplitude of each scattering center is solved in batches by the integral equation model method:

[0040]

[0041] in, The field is the scattering field of a rough surface, where N is the number of rough surface elements, k is the wavenumber, and r is the wavenumber. n Let r be the position vector of the electrically large mesh, r′ be the incident wave vector, and exp(jψ) be the random phase correction term, which needs to be added when the mesh is electrically large. The range of ψ is [0:2π]. n The amplitude of the electrically large grid scattering field is denoted as .

[0042] Furthermore, the step of uniformly refining the mesh in the coupling region of the target mesh specifically includes:

[0043] Solve for the projection of the incident normal vector onto the xy plane, and solve for the angle between it and the out-of-mesh normal vector one by one. When the angle is 0, the mesh is determined to be a coupled region mesh, which is coupled with the background scattered echo.

[0044] Based on the vertex coordinates of the extracted mesh, the electrically large mesh with coupling is uniformly refined.

[0045] Furthermore, the step of calculating the multiple scattering between the target and the background using a semi-analytical four-path model specifically includes:

[0046] Determine whether the meshes in the coupling region are coplanar based on their coordinates, and then perform diversity analysis on the meshes in the coupling region based on whether they are coplanar.

[0047] Solving the coupled scattering field of one grid in each grid set using a four-path model

[0048] The coupled scattering field of this grid set is: in, φ represents the magnitude of all grids in this grid set. i The phase of the i-th coupled mesh;

[0049] The overall coupled scattering field is obtained by superimposing the coupled scattering fields of all grid sets, which is the multiple scattering between the target and the background.

[0050] Furthermore, the calculation method also includes a model validation step, specifically including:

[0051] The scattering field of the composite target was cross-validated with the full-wave data. When the mean RCS error was less than 3dB, the scattering model was valid.

[0052] If the mean RCS error is greater than 3dB, the target edge structure determination and the solution of each scattering component are redone.

[0053] Furthermore, the calculation method also includes the step of calculating a SAR or ISAR image based on the composite scattered echo data.

[0054] Furthermore, the calculation method also includes the step of generating an electrically large grid model of the composite target based on the independent target grid model and the coarse background terrain grid model.

[0055] As can be seen, the composite target scattering calculation method provided in this disclosure performs non-uniform meshing on the composite target model according to the mesh size requirements of different methods, and constructs a method for establishing composite scene echoes based on PO-EEC-IEM-ASC-FPM: large meshing is performed on the target and rough surface, and PO-EEC and IEM-ASC are used for solving; the target mesh with coupling effect is refined, and FPM is used for solving, which minimizes the unknowns of the model; at the same time, when performing coupling solution, based on the characteristic that the scattering amplitude is the same when the normal vector is the same in the PO method, it is proposed to calculate the coupling scattering amplitude of only one refined coupling surface element, and obtain all coupling amplitudes analytically, reducing the number of repeated calculations and significantly improving the calculation efficiency of the total scattering field of the composite scene.

[0056] Compared with existing technologies, the beneficial effects of this disclosure are: ① Introducing the concept of non-uniform mesh-semi-analytical processing into electromagnetic wave scattering modeling, using different mesh sizes for different methods, maximizing the reduction of unknowns while ensuring accuracy; ② Introducing the concept of semi-analytical processing into electromagnetic wave scattering modeling, obtaining the coupled scattering field of the overall target by calculating the coupled scattering amplitude of a surface element; ③ Judging the edge structure based on the target mesh model, considering target surface reflection and edge diffraction using PO and EEC respectively; ④ Proposing a scattering center model based on the integral equation method, parametrically representing the scattering field of the rough surface, and finally obtaining a high-frequency hybrid scattering model that combines the parametric model and the high-frequency model; ⑤ Verifying the correctness of the calculation through model validation and imaging. Attached Figure Description

[0057] Figure 1 This is a flowchart of the calculation method for the composite target scattering model described in this disclosure;

[0058] Figure 2 (a) and (b) are the target model and background model of Example 1, respectively;

[0059] Figure 3 This is a composite scene model of a destroyer and the sea surface, as shown in Example 1.

[0060] Figure 4 To reduce the size of the destroyer target by a factor of 20, the entire target is divided into a single wavelength grid. After determining the edge structure, the center points of all grids and the midpoints of the edge structures are identified.

[0061] Figure 5 To Figure 4 A schematic diagram of the center points and edge midpoints of all grids after the coupling region is densified into a 1 / 5 wavelength grid;

[0062] Figure 6 The RCS of the destroyer-secondary sea surface composite scene model reduced by 20 times was compared with the full-wave results using the method disclosed in this paper;

[0063] Figure 7 SAR imaging results for a destroyer-secondary sea surface composite scene model; Figure 7 (a) SAR results for destroyer targets segmented at 12.5 wavelengths; Figure 7 (b) SAR results for destroyer targets, segmented into 12.5 wavelengths and with the coupled region encrypted to 1 / 5 wavelength;

[0064] Figure 8 (a) and (b) represent the target model and background model of Example 2, respectively;

[0065] Figure 9 This is the tank-desert composite scene model for example two;

[0066] Figure 10 The tank is divided into a 1 / 2 wavelength grid, and the center point of the grid and the midpoint of the edge structure are determined after the edge structure is identified.

[0067] Figure 11 To refine the mesh in the coupled region to the center point and edge structure midpoint of the mesh after 1 / 5 wavelength;

[0068] Figure 12 RCS comparison results of the tank-desert composite scene model using the method disclosed herein and the full-wave results;

[0069] Figure 13 (a) and (b) are ISAR imaging results of the tank-desert composite scene, respectively; Figure 13 (a) ISAR results for the tank target as a whole, divided by 1 / 2 wavelength; Figure 13 (b) ISAR results for tank targets partitioned at 1 / 2 wavelength and with the coupling region encrypted. Detailed Implementation

[0070] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0071] Appendix Figure 1 The flowchart illustrating an exemplary embodiment of the composite target scattering calculation method according to this disclosure includes the following steps:

[0072] Step 1: Using the commercial CAD software CATIA, create a scaled-down geometric model of the independent target and divide it into mesh files; use the Monte Carlo method to create a rough background terrain mesh model.

[0073] As a preferred embodiment, the independent target mesh model and the background terrain mesh model are also imported into FEKO to generate a composite target mesh model (electrical grid) for visually presenting the overall model.

[0074] Step 2, determine the target mesh edge structure:

[0075] Read the target mesh file and determine the vertex coordinates corresponding to all meshes;

[0076] Solve for each edge vector of each grid cell, and use the edge vectors to solve for the out-of-grid normal vector;

[0077] Iterate through all grids that have two identical vertex numbers and define them as adjacent grids;

[0078] Solve for the angle between the outward normal vectors of each pair of adjacent meshes. If the angle satisfies certain conditions, it is determined to be an edge structure. Obtain the outward normal vectors of the edge endpoints and the two lateral face elements of the mesh.

[0079] Step 3: Establish the scattered echo of an independent target using hybrid physical optics and equivalent current method (PO-EEC):

[0080] The scattered echo of a target divided into a mesh model is equivalent to the superposition of the scattering components of all mesh surfaces and the diffraction components generated by all edge structures, as shown in Equation (1).

[0081]

[0082] Where I and J represent the number of grids and the number of edges, respectively.

[0083] The simplified PO contour integral formula is preferred for calculating the surface scattering components, and its expression is as follows.

[0084]

[0085] in Let i be the surface scattering field of the i-th grid. Represents the outside normal vector of the mesh; Indicates the receiving direction of the unit; T is The projected length on the plane of the flat plate, where M is the total number of edges of the flat plate. It is the position vector of the source point. This represents the length and direction of the m-th side of the plate. For tablet and The vertical unit direction. When T = 0, formula (2) simplifies to:

[0086]

[0087] EEC is preferably used to calculate the diffraction components of the edge structure, and its expression is as follows:

[0088]

[0089]

[0090] in, Let j be the diffraction field of the j-th edge. Let η0 be the direction vector at the edge, and η0 be the free space impedance. and These represent the vertical and horizontal polarization directions, respectively. and These are the start and end points of the edge, respectively. Let I be the midpoint of the edge. PTD and M PTDThe diffraction current and diffraction magnetic current at the edge are respectively represented by formulas (6) to (14). The subscripts 1 and 2 represent the upper surface and the lower surface, respectively, and the superscripts PO and PTD represent the surface component and the edge component, respectively.

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100] in, The incident electric field vector; The incident magnetic field vector; k is the wave number of the incident wave; U(x) is the step function. Calculate the equivalent current. With equivalent magnetic current By replacing the terms in equations (7)-(14) β s →π-β s β i →π-β i , To obtain.

[0101] Step 4: Use a surface element model based on the integral equation method to establish the echo of the rough background terrain:

[0102] For rough surface echoes, the Monte Carlo method can be used to generate the rough surface, and then modeling can be performed using a parametric surface model. In terms of parametric modeling, it is preferable to divide the rough surface into electrically large grids with the imaging resolution as the size (if only RCS is calculated and imaging is not performed, the grid size can be larger). Each grid can be regarded as a point scattering center, and the amplitude of each scattering center is solved in batches using the integral equation model method (IEM).

[0103] The scattering field of a rough background using the integral equation method is shown in equation (15):

[0104]

[0105] Where k is the wave number, r n Let r be the position vector of the electrically large mesh, r′ be the incident wave vector, and exp(jψ) be the random phase correction term, which needs to be added when the mesh is electrically large. The range of ψ is [0:2π]. n Let be the amplitude of the electrically large grid scattering field, as shown in Equation (16).

[0106]

[0107] Among them, S n σ is the grid area. IEM (θ i ′) is the scattering coefficient of the rough mesh in the integral equation method, and its expression is shown in formula (17).

[0108]

[0109] Where, θ i ′ is the incident angle defined in the local coordinate system, which is distinct from the global coordinate system, and is represented by the normal vector of the electrical large grid. The z-axis and y-axis can be represented as... Let θ be the incident vector. i ′ represents the angle between the incident vector and the z-axis of the local coordinate system; k = 2π / λ is the wave number; δ is the root mean square height of the rough surface element; the subscript p indicates horizontal polarization (HH) or vertical polarization (VV); W (n) It is the power spectrum of a rough surface. The expression is as follows:

[0110]

[0111]

[0112]

[0113]

[0114] in,

[0115] sq=(μ r ε r -sin 2 θ i ′) 1 / 2 ,sqs=(μ r ε r -sin 2 θ s ′) 1 / 2

[0116] T v =1+R VV ,T vm=1-R VV ,T h =1+R HH

[0117] T hm =1-R HH ,T p =1+R′,T m =1-R′

[0118]

[0119] In the above formula, μ r and ε r These represent the magnetic permeability and dielectric parameters of the rough surface, respectively.

[0120] Step 5: Uniformly refine the mesh located in the coupled scattering region between the target and the background.

[0121] Solve for the incident normal vector Projecting the vectors onto the xy plane and solving for their intersections with the out-of-mesh normal vectors. The included angle is used to determine if the mesh is a coupled region mesh when the included angle is 0.

[0122] To reduce the multiple scattering path error caused by the electrically large mesh, the electrically large mesh is uniformly refined based on the extracted vertex coordinates. The preferred method is as follows:

[0123] Taking a rectangular grid with length and width Lx and width Ly as an example, assuming the center point is at the origin, when the horizontal direction is denser by a factor of 2m and the vertical direction is denser by a factor of 2n, the coordinates of the center point of each grid after densification are: (±(Lx / 2m+(Lx / 4m)×(m-1)),±(Ly / 2n+(Ly / 4n)×(n-1)));

[0124] In this area, the encrypted face number can be arranged from left to right and from top to bottom.

[0125] Step Six: Based on the target mesh model refined in Step Five, calculate the multiple scattering between the target and the background using a semi-analytical four-path model:

[0126] Multiple scattering between the target and the background is equivalent to three components: target-rough surface, rough surface-target, and rough surface-target-rough surface. These, along with single scattering from the target, constitute a four-path system, which performs well at radar incident angles θ = 30°–50°. The expression for the scattering center-background coupled field is as follows:

[0127]

[0128] in, Let be the reflection coefficient of the rough surface, θ be the incident angle in the global coordinate system, and L2, L3, and L4 be the additional paths for the second, third, and fourth paths, respectively.

[0129] L2=2hcosθ, L3=L2, L4=2L2 (21)

[0130] In the formula, h is the height of the surface element; finally, the coupled fields of all scattering centers are superimposed to calculate the scattering between the target and the background.

[0131] As a preferred approach: determine whether the meshes in the coupling region are coplanar based on their coordinates, and then perform diversity analysis on the coupled meshes based on whether they are coplanar.

[0132] The coupled scattering field of one grid in each grid set is solved using formula (20). Where E sc Using formula (2) to solve, the magnitude of all grids in the grid set can be written as: The coupled scattering field of this grid set can be expressed as Where φ i The phase of the i-th coupled mesh;

[0133] The overall coupled scattering field is obtained by superimposing the coupled scattering fields of all grid sets.

[0134] Step 7: Sum the scattering components from steps 3, 4, and 6 to obtain the scattering field of the composite target.

[0135] As a preferred embodiment, the exemplary embodiment further includes the following steps:

[0136] Step 8, Model Validation:

[0137] The obtained composite target scattering field is cross-validated with the full-wave data. When the mean RCS error is less than 3dB, the scattering model is valid. When the mean RCS error is greater than 3dB, return to step two to re-determine the edge structure and solve for each scattering component.

[0138] Step 9: Calculate SAR or ISAR images based on the verified composite scattering echo data.

[0139] Application example:

[0140] Example 1: Destroyer - Sea Surface Model

[0141] Establish a CAD model of the destroyer-sea surface composite target, such as Figure 3 As shown, the ship has a total length of 179.4m, with a triangular bow of 69m, a maximum width of 12m, a maximum height of 26.4m, and a draft of 1m. Figure 2 As shown in (a); the sea surface is a secondary sea surface of 200×200m. Figure 2(b) is a two-scale ocean spectrum formed by the AKFung spectrum, with a wind speed of 5 m / s and a dielectric parameter of 80.7 + i20.7. Additionally, to conserve computational resources for the full-wave method, the overall size of the composite target was reduced by a factor of 20 when comparing RCS.

[0142] When solving for the scattering of the target, the target is divided into a large-area model with an element size of 12.5 wavelengths. The PO method is used to solve the scattering field of each element. The coplanarity of elements is determined by the angle between the normals of adjacent elements. When the angle is greater than 10°, it is considered that there is an edge structure between the two meshes. The EEC method is used to solve the diffraction field. The results of the edge structure determination are as follows: Figure 4 As shown, black dots represent the center points of the target mesh, and red dots represent the midpoints of the edge structures. The sea surface scattering field is solved using an IEM-based scattering center model, with each sea surface element having a size of 3m. The angle between the projections of the incident normal and the element normal onto the xy-plane is used to determine if a mesh belongs to the coupling region; if they are parallel, the mesh is considered a coupling region mesh. The coupling region mesh is then uniformly refined, with the element size after refinement being 1 / 5 of a wavelength. The four-path coupling amplitude of the first mesh is calculated, and this amplitude is defined as the amplitude of all coupling region meshes. The coupled scattering field is then solved in conjunction with the phase of the coupling region meshes. The result after refinement is shown below. Figure 5 As shown, the blue dots represent the center points of the mesh in the encrypted coupling region, and the RCS of the composite scene is as follows: Figure 6 As shown, the mean RCS error between the method described in this disclosure and the full-wave method is 1.77 dB.

[0143] When the parameters of the spaceborne SAR radar are as shown in the table below, the final SAR image results under VV polarization are shown below. Figure 7 (a) and Figure 7 As shown in (b) Figure 7 (a) is a SAR image without considering multiple scattering effects. Figure 7 (b) SAR image considering multiple scattering effect. It can be seen that when considering multiple scattering components, ghosting occurs on the illuminated side. Figure 7 (b) The imaging process takes approximately 3 hours.

[0144]

[0145] Example 2: Tank-Desert Model

[0146] Establish a CAD model of the tank-desert composite target, such as Figure 8 As shown, the tank model has a total length of 7.92m, including a hull length of 6.79m and a gun barrel length of 4.12m. Its total width is 3.56m and its total height is 1.92m, with the hull height being 1.42m. The desert scene is 10m long along the x-axis and 16m long along the y-axis. Its surface roughness parameters are shown in the table below.

[0147]

[0148] The tank target is segmented using a 1 / 2 wavelength, and the edge detection and encryption results are as follows: Figure 10 and Figure 11 As shown, the judgment and encryption results are good; the surface element size of the desert rough surface is 0.75m, and the RCS is as follows: Figure 12 As shown, the mean RCS error between the method disclosed herein and the full-wave method is 0.96 dB. When imaging is performed using the ISAR parameters in the table below, the results are as follows... Figure 13 (a) and (b) show the imaging effects without and with a four-path scattering method, respectively. It can be seen that considering the multiple scattering components, ghosting will occur on the illuminated side. Figure 13 (b) The imaging time is 350s.

[0149]

[0150] In summary, the above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for fast calculation of composite target scattering based on non-uniform grid model, comprising the following steps: obtaining or constructing an independent target grid model and a rough background terrain grid model; judging the edge structure of the target grid; establishing the scattering echo of the independent target by using the hybrid physical optics method and the equivalent current method; establishing the echo of the rough background terrain by using the surface element model based on the integral equation method; uniformly encrypting the coupling area grid in the target grid, and calculating the multiple scattering between the target and the background by using the semi-analytical four-path model; summing the scattering echo of the independent target, the echo of the rough background terrain, and the multiple scattering between the target and the background to obtain the scattering field of the composite target; the step of establishing the scattering echo of the independent target by using the hybrid physical optics method and the equivalent current method, specifically comprising: the scattering echo of the independent target divided into a grid model is equivalent to the superposition of scattering components of all grid surfaces and diffraction components generated by all edge structures: wherein, is the independent target scattering echo, is the surface scattering field of the ith grid, is the diffraction field of the jth edge, I and J represent the number of grids and edges, respectively; calculating the surface scattering component by using the simplified PO contour integral formula: wherein, represents the normal vector outside the grid; represents the unit receiving direction; T is the projection length on the plane of the panel, M is the total number of edges of the panel, is the position vector of the source point, represents the length and direction of the mth edge of the panel, is the unit direction perpendicular to the panel. calculating the diffraction component of the edge structure by using the EEC: where, is the direction vector of the edge, η0is the free space impedance, and represent the vertical and horizontal polarization directions, respectively, and are the start and end points of the edge, respectively, is the midpoint of the edge, I PTD and M PTD are the diffraction current and diffraction magnetic current of the edge, respectively.

2. The computing method of claim 1, wherein, the step of judging the edge structure of the target grid, specifically comprising: reading the target grid file to determine the vertex coordinates corresponding to all grids; solving each edge vector of each grid, and solving the outward normal vector of the grid by using the edge vector; traversing to find all grids with two same grid vertex numbers, which are defined as adjacent grids; solving the included angle of the outward normal vectors of each group of adjacent grids, and determining the edge structure when the included angle meets certain conditions, and obtaining the outward normal vectors of the grid edge endpoints and the two side surface elements.

3. The computing method of claim 1, wherein, the step of establishing the echo of the rough background terrain by using the surface element model based on the integral equation method, specifically comprising: dividing the rough surface into electric large grids with imaging resolution as the size, and regarding each grid as a point scattering center, and solving the amplitude of each scattering center by using the integral equation method model; where, is the rough surface scattering field, N is the number of surface elements, k is the wave number, r n is the position vector of the electrically large grid, r' is the incident wave vector, exp(jψ) is a random phase correction term that needs to be added when the grid is electrically large, ψ is in the range [0:2π], A n is the amplitude of the electrically large grid scattering field.

4. The computing method of claim 2, wherein, the step of uniformly encrypting the coupling area grid in the target grid, specifically comprising: solving the projection of the incident normal vector in the xy plane, and solving the included angle with the outward normal vector one by one, and determining that the grid is a coupling area grid when the included angle is 0, which has coupling with the background scattering echo; uniformly encrypting the electric large grid with coupling according to the vertex coordinates of the extracted grid.

5. The computing method of claim 2, wherein, the step of calculating the multiple scattering between the target and the background by using the semi-analytical four-path model, specifically comprising: judging whether the coupling area grid is coplanar according to the coordinates of the coupling area grid, and dividing the coupling area grid into sets according to whether it is coplanar; Solving the coupled scattering field of one grid in each grid set using a four-path model The coupled scattering field of the grid set is then: where, is the amplitude of the grid set, φ i is the phase of the i-th grid. superimposing the coupling scattering fields of all grid sets to obtain the total coupling scattering field, i.e. the multiple scattering between the target and the background. 6.The calculation method according to any one of claims 1-5, further comprising a model verification step, specifically comprising: verifying the scattering field of the composite target with full-wave data, and when the RCS mean error is less than 3dB, the scattering model is established; when the RCS mean error is greater than 3dB, re-determining the target edge structure and solving each scattering component.

7. The computational method of claim 6, further comprising the step of computing a SAR or ISAR image from the composite scattering echo data.

8. The computational method of claim 1, further comprising the step of generating an electrical large mesh model of a composite target from an independent target mesh model and a coarse background terrain mesh model.

Images