An electromagnetic scattering prediction method for metal-dielectric composite structures based on measurement fusion

Abstract

The present invention discloses an electromagnetic scattering prediction method for metal-dielectric composite structures based on measurement and calculation fusion, belonging to the technical field of electromagnetic scattering characteristics. The electromagnetic characteristics of dielectric materials are characterized by reflection coefficients; the correlation between the reflected field of dielectric materials and the surface reflection coefficient is established, and the reflection coefficient based on the reflected field test data is used to replace the reflection characteristics of dielectric materials; electromagnetic modeling of metal structures is carried out; analysis of direct surface reflection and mutual coupling scattering between structures under the incidence of radar waves is carried out; the test results of reflection coefficients are correlated with optical calculations to determine the data format of measurement and calculation fusion; at each incident angle of radar waves, each sub-aperture surface thread is assigned to different GPU nodes for parallel calculation. Based on the idea of measurement and calculation fusion, the solution is reduced to the vector superposition of electromagnetic scattering fields in different regions, and through the correlation processing of regional boundaries, the collaborative use of testing and calculation is realized, and the accuracy and efficiency of target electromagnetic scattering characteristic prediction are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of electromagnetic scattering characteristics research. More specifically, it relates to an electromagnetic scattering simulation prediction for a composite structure containing dielectric materials, which can be applied to evaluate the electromagnetic scattering characteristics after the target uses radar wave stealth dielectric materials. Background Art

[0002] In terms of scattering characteristics, traditional shape stealth technologies mainly rely on large-angle outward or inward fluttering of the outer surface of the target. Although they play a certain role in controlling and reducing the electromagnetic scattering characteristics of the target, the effect is very limited. At the same time, it also brings problems such as reduced cabin volume utilization and effective load. With the development of radar wave stealth dielectric material technology, the application of dielectric materials on various target platforms has become an inevitable trend. How to predict and control the application effect of radar wave stealth dielectric materials has become one of the key technologies in the engineering design of various target platform models. As an essential quantitative evaluation method for the electromagnetic scattering prediction of metal-dielectric composite structures, by obtaining the change in the electromagnetic scattering characteristics of the target, that is, the radar cross-section, the improvement effect of various stealth measures on the ship's RCS control can be grasped.

[0003] The accuracy of target electromagnetic scattering prediction determines the fidelity of target structure simulation. In theory, the full-wave numerical method has the advantages of complete geometric and electromagnetic parameter characterization and can accurately model the electromagnetic scattering of composite structures. However, when dealing with the scattering problem of large-scale targets in the microwave band, it will be limited by computing resources due to ultra-large-scale grid discretization and is extremely prone to iterative errors. When the full-wave method cannot meet the requirements of engineering applications, currently, the optical calculation method is mainly used to simulate and predict the electromagnetic scattering characteristics of large-scale complex targets. Although this method has high computational efficiency, it is usually applicable to electrically large-scale ideal conductor targets and cannot guarantee the simulation accuracy of materials, restricting the accuracy of electromagnetic scattering prediction. How to give full play to the advantages of the optical approximation algorithm and accurately calculate the influence of dielectric materials is the key problem in predicting the application effect of radar wave stealth dielectric materials.

[0004] Currently, there is still a blank in the electromagnetic scattering prediction method of composite structures using the measurement and calculation fusion method. Therefore, how to realize the electromagnetic scattering prediction method of metal-dielectric composite structures based on measurement and calculation fusion is a technical problem that needs to be solved urgently. Summary of the Invention

[0005] In response to the need to predict and control the application effect of radar wave stealth dielectric materials, an electromagnetic scattering prediction method for metal-dielectric composite structures based on measurement and calculation fusion is proposed. Based on the idea of "measurement and calculation fusion", the solution of the entire problem is attributed to the vector superposition of electromagnetic scattering fields in different regions. Through the associated processing of regional boundaries, the coordinated use of testing and calculation is realized, and the accuracy and efficiency of target electromagnetic scattering characteristic prediction are achieved.

[0006] To achieve the above object, the present invention provides an electromagnetic scattering prediction method for metal-dielectric composite structures based on measurement and calculation fusion, including:

[0007] (1) For non-uniform single-layer dielectrics or multi-layer dielectrics, the electromagnetic characteristics of the dielectric material are characterized by the reflection coefficient;

[0008] (2) Establish the correlation between the reflected field of the dielectric material and the surface reflection coefficient, and use the reflection coefficient based on the reflected field test data to replace the reflection characteristics of the dielectric material;

[0009] (3) Conduct electromagnetic modeling of the metal structure with electromagnetic grid fitting for multi-scale structures, and perform surface element meshing;

[0010] (4) Based on the equivalent electromagnetic current mechanism, analyze the surface direct reflection and inter-structure mutual coupling scattering under the incidence of radar waves;

[0011] (5) Correlate the reflection coefficient test results with optical calculations to obtain the scattering field and determine the measurement and calculation fusion data format;

[0012] (6) At each incident angle of the radar wave, divide the virtual aperture plane into multiple sub-aperture planes, and assign the threads corresponding to each sub-aperture plane to different GPU nodes for parallel calculation.

[0013] In some alternative embodiments, step (1) includes:

[0014] For a single-layer dielectric material, when an electromagnetic wave projects onto the surface of dielectric 0 and dielectric 1, where dielectric 0 is air and dielectric 1 is the dielectric material coated on the metal plate, obtain the TE-wave and TM-wave reflection coefficients of the single-layer dielectric-coated flat plate;

[0015] The reflection coefficient of the multi-layer dielectric material is based on the reflection coefficients of each layer of dielectric material, and the reflection coefficient of the top layer is obtained through the recurrence of the reflection coefficients between each layer.

[0016] In some alternative embodiments, step (2) includes:

[0017] Considering an isotropic homogeneous dielectric, test the surface reflected field under a set of different incident directions. For each incident direction, obtain the scattering coefficient in the corresponding reflection direction according to the amplitude and phase of the reflected field, and use interpolation to obtain the reflection coefficients under different incident directions.

[0018] In some alternative embodiments, step (3) includes:

[0019] For the preprocessing of any complex electromagnetic structure, based on multi-parameter judgment, perform non-uniform grid division according to local particularities;

[0020] For the mesh processing of complex curvature structures, based on the mesh feature region division, surface reconstruction, and non-uniform mesh processing, according to the original surface boundary point information, the extraction of isoparametric line structure control points is adopted, and based on multi-parameter information, the local special deformation mesh processing is established.

[0021] In some alternative embodiments, step (4) includes:

[0022] By Conduct the surface direct reflection and inter-structure mutual coupling scattering analysis under the incident radar wave, where is the total magnetic field of surface reflection, is the total electric field of surface reflection, is the magnetic field of surface direct reflection, is the electric field of surface direct reflection, is the magnetic field of inter-structure mutual coupling scattering, is the electric field of inter-structure mutual coupling scattering.

[0023] In some alternative embodiments, step (5) includes:

[0024] By Obtain the scattering field, where s1 is a closed surface, is the radius vector of the scattered wave, is the outer normal vector of the s1 surface, is the direction perpendicular to the incident plane, is the radius vector of the incident wave, and r′ is the radius vector of the source point.

[0025] In some alternative embodiments, step (6) includes:

[0026] At each incident angle of the radar wave, divide the virtual aperture surface into multiple sub-aperture surfaces, and assign the threads corresponding to each sub-aperture surface to different GPU nodes. After each node calculates the scattering field corresponding to its sub-aperture surface, record its calculation time, and broadcast all the calculation times to each GPU node;

[0027] For the first incident angle, evenly divide the aperture surface, and for subsequent incident angles, dynamically adjust the division of the aperture surface based on the calculation time recorded in the previous angle to achieve load balancing among the nodes.

[0028] In some alternative embodiments, the parallel bounce ray method based on the GPU cluster is implemented by combining two strategies. The first is to assign the calculation tasks corresponding to consecutive angles to each GPU; the second is to divide the virtual aperture surface into multiple sub-aperture surfaces according to the number of GPUs participating in the calculation at each incident angle.

[0029] Generally speaking, compared with the prior art, the above technical solution conceived by the present invention can achieve the following beneficial effects:

[0030] (1) Characterize the electromagnetic properties of dielectric materials using the reflection coefficient, and solve the problem of dielectric material modeling caused by the inability to directly obtain the permittivity and permeability;

[0031] (2) Adopt the method of measurement and fusion to establish the correlation between the reflection field of dielectric materials and the surface reflection coefficient, and use the reflection coefficient based on the reflection field test data to replace the calculation of the reflection characteristics of dielectric materials, so as to solve the problem of difficult direct modeling and calculation of dielectric materials. Description of the Drawings

[0032] Figure 1 is an electromagnetic scattering prediction process of a metal-dielectric composite structure based on measurement and fusion provided by an embodiment of the present invention;

[0033] Figure 2 is a reflection schematic diagram of a single-layer dielectric provided by an embodiment of the present invention;

[0034] Figure 3 is a schematic diagram of a multi-layer dielectric provided by an embodiment of the present invention;

[0035] Figure 4 is a reflection schematic diagram of a plane provided by an embodiment of the present invention;

[0036] Figure 5 is a flow chart of a local special deformation grid processing method provided by an embodiment of the present invention;

[0037] Figure 6 are two parallel strategies of an optical ray calculation method provided by an embodiment of the present invention, wherein (a) is the angle allocation strategy and (b) is the reference plane division strategy;

[0038] Figure 7 is a combined target using stealth dielectric materials provided by an embodiment of the present invention;

[0039] Figure 8 is a test result of the reflection coefficient of a dielectric material provided by an embodiment of the present invention;

[0040] Figure 9 is a test image of the radar cross section of a compact range provided by an embodiment of the present invention;

[0041] Figure 10 is a comparison curve of the radar cross section between the present prediction method and the test result provided by an embodiment of the present invention. Detailed Embodiments

[0042] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0043] As Figure 1 shown, a method for predicting electromagnetic scattering of a metal-dielectric composite structure based on measurement fusion provided by an embodiment of the present invention includes:

[0044] (1) Characterizing the electromagnetic properties of dielectric materials based on reflection coefficients;

[0045] For non-uniform single-layer dielectrics or multi-layer dielectrics, the reflection coefficient is used to characterize the electromagnetic properties of the dielectric material, solving the problem of dielectric material modeling caused by the inability to directly obtain the permittivity and permeability.

[0046] 1) Reflection coefficient of single-layer dielectric material

[0047] As Figure 2 shown, for a single-layer dielectric material, when an electromagnetic wave is incident on the surface of dielectric 0 and dielectric 1, reflection occurs. Dielectric 0 is air, and dielectric 1 is the dielectric material coated on the metal plate.

[0048] The calculation formulas for the TE-wave and TM-wave reflection coefficients of a single-layer dielectric-coated flat plate are shown in Table 1 below:

[0049] Table 1 Reflection coefficient (Γ) of a single-layer dielectric flat plate

[0050]

[0051] Among them, is the incident angle, ω is the angular frequency of the incident wave, ε0 and μ0 are the complex permittivity and complex permeability of free space respectively, ε1 = ε′1 - ∈″1j, μ1 = μ′1 - μ″1j are the complex permittivity and complex permeability of the dielectric region,

[0052] 2) Reflection coefficient of multi-layer dielectric material

[0053] The reflection coefficient of the multi-layer dielectric material is based on the reflection coefficients of each layer of dielectric material. That is, the reflection coefficient of the top layer is obtained through the recursion of the reflection coefficients between each layer. As Figure 3 is a schematic diagram of the multi-layer dielectric material, and the subscript l represents the coefficient of the l-th layer. A l represents the coefficient of all wave components propagating in the +z direction, B l represents the coefficient of all wave components propagating in the -z direction. Then the equivalent reflection coefficient of the multi-layer dielectric material

[0054] The iterative formula for the reflection coefficient is obtained from the following formula:

[0055]

[0056] where k l is the wavenumber of the l-th layer, ε l , μ l are the complex permittivity and complex permeability of the medium region of the l-th layer respectively, and d l is the thickness of the l-th layer.

[0057] (2) Reflection coefficient extraction and interpolation

[0058] Establish the correlation between the reflection field of the dielectric material and the surface reflection coefficient, and use the reflection coefficient based on the reflection field test data to replace the calculation of the reflection characteristics of the dielectric material, so as to solve the problem that it is difficult to directly model and calculate the dielectric material.

[0059] 1) Extraction of surface reflection coefficients at different angles

[0060] As Figure 4 shown, considering an isotropic homogeneous medium, it is necessary to obtain the surface reflection coefficients in the case where the incident angle ranges from 0 degrees to 90 degrees. Therefore, it is necessary to measure a set of surface reflection fields at different incident directions. For each incident direction, the scattering coefficient in the corresponding reflection direction is obtained based on the amplitude and phase of the reflection field.

[0061] The plane where the electromagnetic wave incident direction and the plane normal are located is the incident plane. The angle θ between the electromagnetic wave incident direction and the plane normal is the incident angle of the electromagnetic wave. The angle between the electromagnetic wave reflection direction and the plane normal is the reflection angle, and its magnitude is the same as the incident angle.

[0062] The measured scattering field is a field related to the measurement material area, the measurement position and the distance from the measurement material. The real part and the imaginary part of the field are greatly affected by the phase. Therefore, it is necessary to use a PEC flat plate as a reference to obtain the surface scattering coefficient. The relationship is as follows:

[0063]

[0064] where Γ DIE (θ) is the reflection coefficient of the dielectric material at the incident angle of θ, Γ PEC is the reflection coefficient of the PEC (the reflection coefficient of the PEC for TE waves is -1, and the reflection coefficient for TM waves is 1), are the scattered far fields (including the real part and the imaginary part) of the dielectric material and the PEC in the corresponding reflection directions at the incident angle of θ respectively.

[0065] 2) Reflection coefficient interpolation

[0066] In practical applications, calculations are carried out by obtaining the reflection coefficients at typical incident angles. Therefore, an interpolation algorithm is required to interpolate the known measured reflection coefficients.

[0067] In the embodiments of the present invention, a linear interpolation algorithm is used to obtain the reflection coefficients at different angles. Linear interpolation is a relatively simple interpolation method, and its interpolation function is a first-degree polynomial.

[0068] Assume that the known coordinates are (x0, y0) and (x1, y1). To obtain the value of y at a certain position x within the interval [x0, x1], its calculation formula is:

[0069]

[0070] In addition, assume that the angular range for measuring the reflection coefficient is [a, b] (a > °0, b < 90°). Then it is considered that when the incident angle is within the interval [0°, a), its reflection coefficient is the same as that when the incident angle is a; when the incident angle is within the range (b, 90°], its reflection coefficient is the same as that when the incident angle is b.

[0071] (3) Electromagnetic modeling of metal structures

[0072] The electromagnetic modeling of metal structures mainly focuses on electromagnetic grid fitting for multi-scale structures. The surface element meshing is not only used to simulate the target geometric surface or boundary, but also the quality of the meshing directly determines the behavior of the surface element currents and their mutual interactions. As the resolution unit, the regularity of the surface element shape plays a decisive role in controlling the error of the entire iterative solution.

[0073] For the preprocessing of any complex electromagnetic structure, the present invention is based on the judgment of multiple parameters (such as dielectric constant and curvature), and non-uniform grid division is carried out according to local particularities, which ensures the calculation accuracy while maximizing the speed.

[0074] Regarding the grid processing problem of complex curvature structures, based on the research of grid feature region division, surface reconstruction, and non-uniform grid processing methods, based on the information of the original surface boundary points, an isoparametric line structure control point extraction method is adopted, and based on the information of multiple parameters (such as dielectric constant and curvature), a local particularity deformation grid processing method (such as Figure 5 ) is established to solve the problem of algorithm adaptability of multi-scale structure grids.

[0075] (4) Optical calculation of metal structures

[0076] When a metallic target is irradiated by radar waves, the energy incident on the object will be scattered in all directions. The scattered field mainly includes single reflections and multiple scatterings generated on the surface of the object due to sudden changes in wave impedance. That is, under the excitation of the incident wave, the scattered field of the target is a combination of single scattering of the structure and scattering due to the interaction between structures. Based on the analysis of the induced current mechanism, the scattered field is the secondary radiation from the induced electromagnetic current caused by the incident wave on the surface of the object. The spatial distribution of the scattered field is called the scattering pattern, which is related to the size, shape, and structure of the object, as well as the frequency and polarization of the incident wave. Based on the equivalent electromagnetic current mechanism, the analysis of surface direct reflection and inter-structure mutual coupling scattering under radar wave incidence is carried out.

[0077]

[0078]

[0079] Among them, are the total magnetic field and total electric field of the surface reflection respectively, are the magnetic field and electric field of the surface direct reflection respectively, are the magnetic field and electric field of the inter-structure mutual coupling scattering respectively.

[0080] According to the combination of vector Green's theorem and Maxwell's equations, the integral equation of the single scattering field is obtained as follows:

[0081]

[0082]

[0083] Among them, is the distance between the scattering observation point and the source point, is the unit outer normal vector on the surface S, k is the wave number, and μ is the magnetic permeability of the medium.

[0084] (5) Measurement and fusion

[0085] 1) Correlation between the test results of the reflection coefficient and the optical calculation method

[0086] In electromagnetic wave theory, the definitions of the induced current and induced magnetic current on the target surface are as follows: and Among them, and are the total electric field and magnetic field on the surface. Considering the far-field condition, the incident field is generally expressed in spherical coordinates. Assuming that the incident electromagnetic wave propagates along direction, the electric field and the magnetic field are in the plane. Among them, is perpendicular to the incident plane, Parallel to the incident plane. The induced current and induced magnetic current are in the following forms:

[0087]

[0088]

[0089] Where Γ ⊥ and Γ / / are the reflection coefficients of TE wave and TM wave respectively, E θ and E φ are the electric field components of plane and plane respectively, and H θ and H φ are the magnetic field components of plane and plane respectively.

[0090] The Stratton-Chu integral formula is as follows:

[0091]

[0092] Where g is the Green's function and ε is the permittivity of the medium.

[0093] Under the condition of an ideal conductor, the incident wave is completely reflected by the boundary, and the tangential electric field and the normal magnetic field are continuous on the conductor surface. The Stratton-Chu integral formula is simplified to:

[0094]

[0095] Where is the induced current on the target surface.

[0096] Under the far-field condition, the gradient of the Green's function can be simplified to Where is the unit vector direction from the field point to the source point, which can be approximated as the scattering direction, Substituting the gradient of the Green's function into the magnetic field integral formula, we get:

[0097]

[0098] From duality (H→-E / η, J→M / η), the electric field integral equation can be obtained:

[0099]

[0100] Where is the induced magnetic current on the target surface.

[0101] The mixed electric field integral equation and magnetic field integral equation:

[0102]

[0103]

[0104] Among them, η is the wave impedance of free space, is the outer normal of the S surface.

[0105]

[0106] Assume the incident wave is

[0107]

[0108] Among them, E0 is the amplitude of the incident wave, is the wave vector, is the radius vector.

[0109] Let be the direction perpendicular to the incident plane, be the direction parallel to the incident plane. The incident wave can be decomposed into TE wave and TM wave, and can be expressed as:

[0110]

[0111]

[0112]

[0113] Among them is the incident electric field.

[0114] Let Γ ⊥ = Γ′ ⊥ + Γ″ ⊥ j be the reflection coefficient of the TE wave, Γ || = Γ′ ‖ + Γ″ ‖ j be the reflection coefficient of the TM wave, then the reflected wave can be expressed as:

[0115]

[0116] The finally output scattered field is:

[0117]

[0118] Among them, s1 is the closed surface, is the radius vector of the scattered wave, is the outer normal vector of the S1 surface, is the direction perpendicular to the incident plane, is the radius vector of the incident wave, r′ is the radius vector of the source point.

[0119] For an isotropic and homogeneous medium, its reflection coefficient is related to the incident wave frequency, incident wave polarization, incident angle, and the thickness of the medium material.

[0120] 2) Measurement and calculation fusion data format

[0121] In the research on the semi-physical simulation method based on measurement and calculation fusion, the required data types and requirements are shown in Table 2 below:

[0122] Table 2 Measurement and calculation fusion data format

[0123]

[0124] Note: The incident plane refers to the plane where the incident wave direction and the normal vector of the measurement flat plate are located.

[0125] (6) Parallel computing

[0126] At each incident angle of the radar wave, the virtual aperture surface is divided into multiple sub-aperture surfaces, and the threads corresponding to each sub-aperture surface are assigned to different GPU nodes. After each node calculates the scattering field corresponding to its sub-aperture surface using the GPU-based optical ray calculation method, it will record its calculation time. And broadcast all the above calculation times to each GPU node. For the first incident angle, the aperture surface is evenly divided, and for subsequent incident angles, the division of the aperture surface is dynamically adjusted based on the calculation time recorded at the previous angle in order to achieve load balancing among nodes.

[0127] The parallel bouncing ray method based on the GPU cluster is implemented by combining two strategies, as Figure 6 shown, where (a) is the angle assignment strategy and (b) is the reference surface division strategy. The first is to assign the calculation tasks corresponding to consecutive angles to each GPU (angle assignment strategy). The second is to divide the virtual aperture surface into multiple sub-aperture surfaces according to the number of GPUs participating in the calculation at each incident angle (aperture surface division strategy). For the first parallel strategy, based on the case where the calculation loads of adjacent angles are the same, the number of angles to be calculated is not necessarily an exact integer multiple of the number of GPUs. To avoid the situation where some GPUs are fully loaded while others are idle, the aperture surface division strategy is still adopted for the remaining angles after equal division of the GPUs.

[0128] For a composite target using stealth medium materials, such as Figure 7 , with a maximum longitudinal dimension of 1 meter and a maximum transverse dimension of 0.4 meter. The top surface is coated with wave-absorbing material, close to the metal surface. The test data of the reflection coefficient at a frequency of 10 GHz is as Figure 8 . The simulation prediction uses the measurement and calculation fusion method proposed in the present invention and is compared with the anechoic chamber test as Figure 9 . The comparison results are as Figure 10 .

[0129] Simulation frequency: 10.5 GHz;

[0130] Polarization modes: VV, HH;

[0131] Horizontal incident azimuth angle: -180° to 180°, with an angular step of 0.5 degrees;

[0132] Calculation parameter: Monostatic radar cross section.

[0133] The hardware platform for prediction calculation is a 64-core server with 32 GB of memory

[0134] It should be noted that according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, or two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.

[0135] Those skilled in the art can easily understand that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An electromagnetic scattering prediction method for metal-dielectric composite structures based on measurement and fusion, characterized in that Including: (1) For non-uniform single-layer or multi-layer media, the electromagnetic characteristics of the media materials are characterized by the reflection coefficient. (2) The correlation between the reflection field of the media material and the surface reflection coefficient is established, and the reflection coefficient based on the reflection field test data is used to replace the reflection characteristics of the media material. (3) Electromagnetic modeling of metal structures with electromagnetic grid fitting for multi-scale structures is carried out, and surface element meshing is performed. (4) Based on the equivalent electromagnetic current mechanism, the surface direct reflection and inter-structure mutual coupling scattering analysis under radar wave incidence are carried out. (5) The test results of the reflection coefficient are correlated with the optical calculation to obtain the scattering field, and the measurement fusion data format is determined. (6) At each incident angle of the radar wave, the virtual aperture surface is divided into multiple sub-aperture surfaces, and the threads corresponding to each sub-aperture surface are assigned to different GPU nodes for parallel calculation. Step (5) includes: From the scattered field is obtained, where s1 is a closed surface, is the radius vector of the scattered wave, is the outer normal vector of the s1 surface, is the direction perpendicular to the incident plane, is the radius vector of the incident wave, and r′ is the radius vector of the source point.

2. The method according to claim 1, wherein Step (1) includes: For single-layer dielectric materials, when electromagnetic waves are projected onto the surfaces of dielectric 0 and dielectric 1, where dielectric 0 is air and dielectric 1 is the dielectric material coated on the metal plate, the TE-wave and TM-wave reflection coefficients of the single-layer dielectric-coated plate are obtained. The reflection coefficient of multi-layer dielectric materials is based on the reflection coefficients of each layer of dielectric materials, and the reflection coefficient of the top layer is obtained through the recursion of the reflection coefficients between each layer.

3. The method according to claim 2, wherein Step (2) includes: Considering isotropic homogeneous media, the surface reflection fields at a group of different incident directions are measured. For each incident direction, the scattering coefficient in the corresponding reflection direction is obtained according to the amplitude and phase of the reflection field, and the reflection coefficients at different incident directions are obtained using interpolation.

4. The method according to claim 3, characterized in that, Step (3) includes: For the preprocessing of any complex electromagnetic structure, based on multi-parameter judgment, non-uniform grid division is carried out according to local particularities. For the grid processing of complex curvature structures, based on grid feature region division, surface reconstruction and non-uniform grid processing, according to the original surface boundary point information, isoparametric line structure control points are extracted, and local particularity deformation grid processing is established based on multi-parameter information.

5. The method according to claim 4, wherein Step (4) includes: By conduct surface direct reflection and inter-structure mutual coupling scattering analysis under radar wave incidence, where is the total magnetic field of surface reflection, is the total electric field of surface reflection, is the magnetic field of surface direct reflection, is the electric field of surface direct reflection, is the magnetic field of inter-structure mutual coupling scattering, is the electric field of inter-structure mutual coupling scattering.

6. The method according to claim 5, characterized in that, Step (6) includes: At each incident angle of the radar wave, the virtual aperture surface is divided into multiple sub-aperture surfaces, and the threads corresponding to each sub-aperture surface are assigned to different GPU nodes. After each node calculates the scattering field corresponding to its sub-aperture surface, its calculation time is recorded, and all the calculation times are broadcast to each GPU node. For the first incident angle, the aperture surface is evenly divided, and for subsequent incident angles, the division of the aperture surface is dynamically adjusted based on the calculation time recorded at the previous angle in order to achieve load balancing among the nodes.

7. The method according to claim 6, wherein The parallel shooting and bouncing ray method based on the GPU cluster is implemented in a way that combines two strategies. The first is to assign the calculation tasks corresponding to consecutive angles to each GPU; the second is to divide the virtual aperture surface into multiple sub-aperture surfaces according to the number of GPUs participating in the calculation at each incident angle.

Images