CN101430376B - Target radar scattering cross-section pre-estimation system with graphics electromagnetic computation accelerated by index information - Google Patents

Abstract

The invention discloses a pre-evaluation system for accelerating computation of target radar scattering cross section by graphical electromagnetic computation through index information. The system comprises a surface element modeling module (1), an index information compilation module (2), an entity display module (3), an index information unscrambling module (4), a geometric information analysis module (5) and a target radar scattering cross section acquisition module (6). The surface elements of a target model are numbered, an index list of the surface elements is established, the surface element index and the surface element color are subject to one to one correspondence, the background color is set as black, components of three primary colors namely red, green and blue are zero, the surface element blanking is finished by a GPU, the corresponding surface element index is obtained by the color of pixels so as to obtain normal vectors and depth values of the pixels. The whole process needs no illumination, the surface normal vector is pre-obtained in a drawing process of the surface elements, a color value is read only for once, and twice memory exchange is executed, which is one computation less than the normal graphical electromagnetic computation, thus shortening the computation time.

Description

Utilizing Index Information to Accelerate Graphical Electromagnetic Calculation Target Radar Cross Section Pre-evaluation System

technical field

The present invention relates to a radar cross-section pre-evaluation in stealth design, more particularly, a pre-evaluation system that utilizes index information to accelerate graphic electromagnetic calculation of target radar cross-section.

Background technique

In estimating the radar cross section (RCS, Radar Cross Section) of electrically large and complex targets, because the numerical method is still weak for the RCS calculation of electrically large and complex targets, the graphical electromagnetic calculation method has been paid more and more attention. Compared with the numerical method There is incomparable computing speed.

The graphics electromagnetic calculation method utilizes the graphics acceleration function of the computer display hardware to obtain the normal vector corresponding to each pixel on the target surface through the irradiation of red, green and blue monochromatic light, and uses the depth buffer (Z-Buffer ) to obtain the depth information corresponding to each pixel on the target surface, calculate the phase relationship between the scattering elements, and combine the physical optics approximation and physical diffraction theory of the complex target in the high frequency region to obtain the corresponding pixel on the target surface The scattered field and diffraction field at the pixel point, and the total scattered field of the target are obtained by the coherent superposition of the scattered field at these discrete points. In the above process, the computer GPU is responsible for the calculation of the normal vector of the target surface, and the occlusion between the targets, that is, the depth buffer information, is also completed by the GPU.

Since the graphical electromagnetic calculation method needs two illuminations to obtain the normal vector of the target surface, it involves a lot of time-consuming memory exchanges on the computer, and the GPU is used for graphics floating-point calculations, the accuracy is not as good as that of the CPU, resulting in the target radar cross-section pre-evaluation system. When the operation is extended, the accuracy decreases.

Contents of the invention

The invention utilizes the index information to identify the surface element that has not been blanked, and calculates the scattering integral of the surface element through the computer CPU, so as to obtain the radar scattering cross section of the target.

The target radar cross section pre-evaluation system of the present invention includes a surface element modeling module 1, an index information module 2, an entity display module 3, an interpretation index information module 4, a geometric information analysis module 5 and a target radar cross section acquisition module 6;

Surface element modeling module 1 is used to carry out parametric modeling on the surface of any entity (entity can be special aircraft, ship, vehicle, etc.) using NURBS method to form an entity parametric surface; on the other hand, the entity The parametric surface is meshed to form a surface split model structure, and it is saved in *.dxf file format.

The indexing information module 2 enables each surfel in the split model structure to have a unique color representation through the mapping relationship between the bin index N and the bin color CN(R, G, B).

In the present invention, the mapping relationship between the bin index N and the bin color CN(R, G, B) is N=R·65536+G·256+B.

The entity display module 3 colors the surface split model structure according to the panel color CN(R, G, B); then, uses OpenGL to perform blanking processing on the face split model structure to obtain the colored surface of the unoccluded part of the face split model structure Element F _N (N ∈ {1, 2, 3, ..., m}), N represents the index of the surface element, and m represents the number of surface elements.

Interpreting the index information module 4 first uses OpenGL to read the color in the colored panel F _N (N ∈ {1, 2, 3, ..., m}), and then adopts the mapping relationship N=R·65536+G·256+ B obtains the index N(F _N ) of the colored panel F _N (N∈{1, 2, 3, ..., m}).

The geometric information analysis module 5 accesses the corresponding surfel in the surface split model structure according to the index N(F _N ), and obtains _the geometric information

进行散射积分计算处理，得到目标雷达散射截面。Target radar cross-section acquisition module 6 adopts geometric optics theory (GO), physical optics approximation (PO) and physical diffraction theory (PTD) to analyze geometric information

Scattering integral calculation processing is carried out to obtain the target radar scattering cross section.

Description of drawings

Fig. 1 is a structural block diagram of a target radar cross-section pre-evaluation system of the present invention.

Figure 2 is a simplified diagram of the radar line of sight in the target radar coordinate system.

Detailed ways

The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments.

The target radar scattering cross-section pre-evaluation system of the present invention, on the basis of the graphic electromagnetic method, reduces the number of time-consuming memory exchanges, and the calculation of the normal vector of the target surface is completed by the computer CPU, and at the same time, the single instruction multiple data of the mainstream CPU is used (SINGLE INSTRUCTION MULTIPLE DATA, SIMD) The parallel operation function of the stream processing model accelerates the calculation process of the normal vector. The main technical solution is: number the surface elements of the target model, establish a surface element index list, correspond the surface element index to the surface element color one by one, and set the background color to black, that is, red (R) and green (G) The components of the three primary colors of , blue (B) are all zero, use the GPU to complete the bin blanking, and obtain the corresponding bin index from the color of the pixel, so as to obtain the normal vector and depth value of the pixel. The whole process does not require lighting, and the surface normal vector has been pre-calculated in the surface element drawing stage. It only needs to read the color value once and perform two memory exchanges, which is one less than the general electromagnetic calculation of graphics, thus shortening the calculation time.

Referring to Fig. 1, the target radar cross section pre-evaluation system of the present invention includes a surface element modeling module 1, an index information module 2, an entity display module 3, an interpretation index information module 4, a geometric information analysis module 5 and a target radar Scattering section acquisition module 6;

Surface element modeling module 1 is used to carry out parametric modeling on the surface of any entity (entity can be special aircraft, ship, vehicle, etc.) using NURBS method to form an entity parametric surface; on the other hand, the entity The parametric surface is meshed to form a surface split model structure, and it is saved in *.dxf file format.

The indexing information module 2 enables each surfel in the split model structure to have a unique color representation through the mapping relationship between the bin index N and the bin color CN(R, G, B).

In the present invention, the mapping relationship between the bin index N and the bin color CN(R, G, B) is N=R·65536+G·256+B, where R represents red, G represents green, and B represents blue.

The entity display module 3 colors the surface split model structure according to the panel color CN(R, G, B); then, uses OpenGL to perform blanking processing on the face split model structure to obtain the colored surface of the unoccluded part of the face split model structure Element F _N (N ∈ {1, 2, 3, ..., m}), N represents the index of the surface element, and m represents the number of surface elements.

Interpreting the index information module 4 first uses OpenGL to read the color in the colored panel F _N (N ∈ {1, 2, 3, ..., m}), and then adopts the mapping relationship N=R·65536+G·256+ B obtains the index N(F _N ) of the colored panel F _N (N∈{1, 2, 3, ..., m}).

The geometric information analysis module 5 accesses the corresponding surfel in the surface split model structure according to the index N(F _N ), and obtains _the geometric information

进行散射积分计算处理，得到目标雷达散射截面。Target radar cross-section acquisition module 6 adopts geometric optics theory (GO), physical optics approximation (PO) and physical diffraction theory (PTD) to analyze geometric information

Scattering integral calculation processing is carried out to obtain the target radar scattering cross section.

Referring to shown in Fig. 1, in the present invention, the panel index N in the indexing information module 2 refers to the sequence number needed to identify each panel in the surface split model structure, and the sequence number can be written arbitrarily according to the designer, such as by Numbered sequentially, or alphabetically numbered, etc.

Referring to shown in Fig. 1, in the present invention in the target radar scattering cross-section acquisition module 6, in order to calculate the target radar scattering cross-section under different radar line-of-sight angles, the target radar coordinate system is marked as xyz, and the unit direction vector of the radar line-of-sight is in the target radar coordinate The system can be expressed as (D _x , D _y , D _z ), as shown in Fig. 2 . The target radar needs to rotate around the x-axis and y-axis so that the oz-axis of the target radar coordinate system coincides with the radar line of sight direction. At this time, it is necessary to perform coordinate rotation transformation on the normal vector of the target surface element. The included angle between the target (or target radar) and the coordinate plane oxz is α, and the included angle between the projection of the target on the coordinate plane oxz and the coordinate axis oz is β.

The coordinate transformation matrix R _y [β] after the target radar is rotated counterclockwise around the y axis (according to the right-hand rule) by an angle of β is:

${\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; Y</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; \&beta;</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; =</span>\left[\begin{array}{ccc}\mathrm{<span\; class="notranslate">\; cos</span>}\mathrm{<span\; class="notranslate">\; \&beta;</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; sin</span>}\mathrm{<span\; class="notranslate">\; \&beta;</span>}\\ \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; 1</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}\\ \mathrm{<span\; class="notranslate">\; sin</span>}\mathrm{<span\; class="notranslate">\; \&beta;</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; cos</span>}\mathrm{<span\; class="notranslate">\; \&beta;</span>}\end{array}\right]<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

The coordinate transformation matrix R _X [α] after the target radar rotates clockwise around the ox axis by an angle of α is:

${\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; =</span>\left[\begin{array}{ccc}\mathrm{<span\; class="notranslate">\; 1</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; 0</span>}\\ \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; cos</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}& <span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; sin</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}\\ \mathrm{<span\; class="notranslate">\; 0</span>}& \mathrm{<span\; class="notranslate">\; sin</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}& \mathrm{<span\; class="notranslate">\; cos</span>}\mathrm{<span\; class="notranslate">\; \&alpha;</span>}\end{array}\right]<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

The target radar first rotates around the y-axis and then around the ox-axis twice, that is, the transformation matrix R _transform from the radar sight angle to the target radar coordinate system is:

R _transform ＝R _Y [β]·R _X [α] (3)

To convert the normal vector in the target radar coordinate system to the radar line of sight angle, you only need to multiply the normal vector by the conversion matrix between the two coordinate systems to get it, only involving the basic addition, subtraction, multiplication and square root operation. Therefore, for the normal vector transformation of multiple surface elements, the parallel operation function of the single instruction multiple data (SINGLE INSTRUCTION MULTIPLE DATA, SIMD) stream processing model of the mainstream CPU can be used to accelerate.

Using index information to accelerate graphic electromagnetic calculation target radar scattering cross section pre-evaluation system proposed by the present invention to obtain target radar scattering cross section has the following advantages:

1. The number of memory swaps is small. Generally, the GRECO method needs two illuminations to obtain the target surface normal vector, which involves three time-consuming memory exchanges. This method does not require additional illumination, and requires two memory exchanges to obtain accurate pixel normal vectors and depth values.

2. High precision. Generally, the GRECO method uses the computer GPU to perform illumination calculations to obtain pixel normal vectors. In this method, the pixel normal vectors are calculated by the CPU when the surface element model is established. Since the floating-point calculation accuracy of the GPU is not as good as that of the CPU, this method can obtain accurate Pixel normal vector.

3. The RCS calculation process is accelerated. This method uses the parallel operation function of the single instruction multiple data (SINGLE INSTRUCTION MULTIPLE DATA, SIMD) instruction set of the mainstream CPU to accelerate the rotation transformation of the normal vector, and can also perform parallel operations on the pixel summation in the RCS calculation process.

4. The panel index can be used to access the array storing the panel information, and the accurate panel normal vector and depth value relative to the viewing direction can be obtained. Compared with the floating-point calculation of the normal vector and depth of the computer GPU, the precision is higher.

5. The whole process does not require monochromatic lighting. In order to obtain the surface element normal vector, it only needs to read the color value once and perform a memory exchange, which is twice less than the general electromagnetic calculation of graphics; it is proposed to use the single instruction multiple data (SINGLE INSTRUCTION MULTIPLE DATA) of the mainstream CPU , SIMD) stream processing model's parallel operation function simultaneously processes the calculation of multiple surface element normal vectors and depth values, which improves the calculation speed and shortens the calculation time to a certain extent under the premise of ensuring accuracy.

Claims (3)

1. one kind is utilized the index information target radar scattering cross-section pre-estimation system with graphics electromagnetic computation accelerated, it is characterized in that: include bin MBM (1), produce index information module (2), entity display module (3), understand index information module (4), geological information parsing module (5) and target radar scattering cross-section acquisition module (6);

Bin MBM (1) is used for adopting the NURBS method to carry out parametric modeling to any solid object surface on the one hand, forms entity parametrization surface; On the other hand mesh generation formation face is carried out on this entity parametrization surface and split model structure, and preserve with * .dxf file layout;

(mapping relations B) make face split that each bin has unique color showing in the model structure to produce index information module (2) for R, G by bin index N and bin color CN;

(mapping relations B) are N=R65536+G256+B for R, G for bin index N and bin color CN;

(B) split model structure and paint for R, G by the opposite according to bin color CN for entity display module (3); Then, use the OpenGL opposite to split model structure and carry out elimination of hidden, acquisition face is split coloured bin F of the part that is not blocked in the model structure _N(N ∈ 1,2,3 ..., m}), N represents the bin index, m represents the number of bin;

Understanding index information module (4) at first uses OpenGL to reading coloured bin F _N(N ∈ 1,2,3 ..., the m}) color in adopts mapping relations N=R65536+G256+B to obtain coloured bin F then _N(N ∈ 1,2,3 ..., index N (F m}) _N);

Geological information parsing module (5) is according to index N (F _N) access plane splits corresponding with it bin in the model structure, obtains coloured bin F _N(N ∈ 1,2,3 ..., geological information m})

Target radar scattering cross-section acquisition module (6) adopts theory of geometric optics (GO), physical optics approximate (PO) and physics diffraction theory (PTD) to geological information

Carry out the scattering integral computing, obtain target radar scattering cross-section.

2. the index information target radar scattering cross-section pre-estimation system with graphics electromagnetic computation accelerated that utilizes according to claim 1 is characterized in that: the bin index N in the produce index information module (2) is meant that face splits in the model structure the needed sequence number of each bin of identification.

3. the index information target radar scattering cross-section pre-estimation system with graphics electromagnetic computation accelerated that utilizes according to claim 1, it is characterized in that: in the target radar scattering cross-section acquisition module (6), target makes the oz axle of target-based coordinate system overlap with the radar line of sight direction around x axle and the rotation of y axle; At this moment need the target panel method is vowed that doing rotational transform is R _Transform=R _Y[β] R _X[α], wherein, $R_Y\left[\mathrm{\&beta;}\right]=\left[\begin{array}{ccc}\cos \mathrm{\&beta;}& 0& -\sin \mathrm{\&beta;}\\ 0& 1& 0\\ \sin \mathrm{\&beta;}& 0& \cos \mathrm{\&beta;}\end{array}\right],$ $R_X\left[\mathrm{\&alpha;}\right]=\left[\begin{array}{ccc}1& 0& 0\\ 0& \cos \mathrm{\&alpha;}& -\sin \mathrm{\&alpha;}\\ 0& \sin \mathrm{\&alpha;}& \cos \mathrm{\&alpha;}\end{array}\right],$ α represents the angle of target radar and coordinate plane oxz, and β is illustrated in projection on the coordinate plane oxz and the angle of coordinate axis oz.

Images