Near-field pedestrian RCS equivalent infinitesimal simulation model

Abstract

The invention belongs to the technical field of radar target scattering characteristic modeling, and the method comprises the steps: dispersing a human body surface into an equal-area equivalent infinitesimal set, and solidifying and storing a center coordinate, an outer normal, an area, a region identifier and a material identifier by infinitesimal elements; the material electromagnetic parameters are obtained by indexing the material identifiers in the parameter table. The visible set participates in compound scattering field synthesis, and the shielding set outputs a shielding ratio and shielding scattering intensity statistical magnitude and participates in error evaluation and correction of a synthesis result according to a preset compensation rule. And introducing standard body calibration, fitting in a unified frequency band and distance range to obtain a near-field compensation coefficient kc, and performing magnitude consistency compensation on a synthesis result. And finally, a three-dimensional RCS matrix changing along with the distance, the orientation and the pitching is output and can be used for pedestrian target modeling, radar system simulation and algorithm training, and the consistency, reproducibility and engineering applicability of near-field simulation are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar target scattering characteristics modeling, and more particularly, to a near-field pedestrian RCS equivalent micro-element simulation model. BACKGROUND

[0002] With the rapid development of intelligent driving technology and vehicle-mounted millimeter wave radar systems, pedestrian target detection, identification and safety warning have become one of the key technical problems in the field of road traffic safety. Radar cross section (RCS) as an important physical quantity representing the scattering ability of a target to radar waves directly affects the detection distance, detection probability of the radar to the pedestrian target, and the performance of subsequent target identification and tracking algorithms. Therefore, accurately obtaining the RCS characteristics of pedestrians in actual working scenarios is of great significance for the design, testing and evaluation of vehicle-mounted millimeter wave radar systems.

[0003] In the prior art, the research on pedestrian RCS mainly focuses on two methods: one is based on measurement, which obtains the RCS characteristics of pedestrians by measuring the radar of pedestrian targets in anechoic chambers or outdoor environments; the other is a modeling method based on electromagnetic simulation, which analyzes the scattering characteristics of pedestrians by constructing a pedestrian model and using numerical calculation or approximation methods. However, these existing technologies still have obvious deficiencies in the near-field application scenarios of vehicle-mounted millimeter wave radars.

[0004] Firstly, from the measurement point of view, the measurement process of pedestrian RCS usually has high requirements for the test environment, test equipment and measured object posture, and the test cost is high and the repeatability is poor. At the same time, the measured data often depends on the specific human posture, clothing type and measurement position, and it is difficult to form a unified and stable RCS reference data, and it is difficult to meet the demand of large-scale parameterized simulation in engineering application.

[0005] Secondly, from the simulation modeling point of view, the existing pedestrian RCS simulation methods mostly use simplified far-field assumption or approximate the human body as several regular geometric bodies, and only perform RCS superposition at the overall level. Although this kind of method has lower computational complexity, it does not fully consider the scattering near-field conditions when vehicle-mounted millimeter wave radars detect pedestrians at close range, and it does not organize geometric, material and occlusion information in a unified scattering unit in the model, which makes it difficult to consistently depict the influence of incident spherical wave phase relationship, human self-occlusion and clothing material difference on scattering results, resulting in obvious deviation of simulation results under near-field conditions.

[0006] Therefore, there is an urgent need in the art for a pedestrian RCS simulation model that can solve the contradiction between near-field scattering modeling accuracy and complexity, and comprehensively consider various factors in a unified framework. SUMMARY

[0007] In order to overcome the above-mentioned defects of the prior art, the present application provides a near-field pedestrian RCS equivalent micro-element simulation model, which is obtained by discretizing the surface of a pedestrian into equivalent micro-elements with equal areas and performing scattering calculation, material correction, occlusion visibility determination and near-field compensation in the micro-element layer to solve the problems raised in the above background art.

[0008] The "equivalent micro-element" referred to herein refers to a curved surface discrete micro-element satisfying the equal area rule, and the shape of the micro-element is not limited; in a preferred implementation, a circular flat micro-element is used to obtain a simplified approximate expression and higher calculation efficiency.

[0009] In order to achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0010] A near-field pedestrian RCS equivalent micro-element simulation model, which takes a set of equivalent micro-elements with equal areas as a unified near-field scattering calculation unit and a unified data carrier, limits pedestrian geometry, material, occlusion and compensation to the same micro-element layer to complete closed-loop solving, and avoids parameter inconsistency caused by module concatenation; each micro-element in the micro-element set solidifies static information; radar line direction vectors and distances are calculated in real time from radar positions and micro-element center coordinates, and a reproducible three-dimensional RCS data matrix is directly outputted accordingly; the model is generated and outputted by the following steps: step one, establishing a parameterized geometric body of a pedestrian divides the head, torso, upper limbs and lower limbs into curved surface regions; step two, equal-area equivalent micro-elements are respectively used in each curved surface region to generate micro-elements in each curved surface according to the equal area rule and write into the parameter field; step three, under the given conditions of working frequency and radar spatial position, near-field spherical wave physical optics is used to calculate scattering echoes of micro-elements, and reflection correction is performed according to material identification; step four, clustering is performed according to the direction of micro-elements to the radar direction, only the nearest micro-element to the radar is kept as a visible micro-element in each cluster, and the rest of the micro-elements perform occlusion processing, and the clustering angle threshold is preset according to distance segmentation; step five, the near-field compensation coefficient obtained by standard body calibration at the same frequency and distance is used to uniformly compensate the echoes of visible micro-elements and synthesize the target RCS; step six, the radar position is scanned on the distance of 1 meter to 20 meters, the azimuth angle of 0° to 360° and multiple height slice grids, and a three-dimensional RCS matrix is outputted according to the distance, azimuth angle and height index.

[0011] As a further scheme of the present application, the static information of the micro-element at least includes micro-element center coordinates, micro-element outer normal, micro-element area, micro-element equivalent size parameter, region identification, material identification and occlusion state; the material identification is used to index the corresponding material electromagnetic parameters in the material parameter table.

[0012] As a further scheme of the present application, the pedestrian parameterized geometry of step one is composed of six regions, including head, torso, left arm, right arm, left leg and right leg; wherein the head is described by a sphere, and the torso and limbs are described by a cylinder.

[0013] As a further scheme of the present application, step two includes adopting ring belt division in the spherical region, so that each ring belt has the same area, and the ring belt is segmented in the circumferential direction to generate micro elements satisfying the same area rule; the cylindrical side region is segmented in the height direction and the circumferential direction to generate micro elements satisfying the same area rule; the center of the micro element is taken as the centroid of the corresponding curved surface facet, and the outer normal of the micro element is taken as the outer normal of the curved surface at the centroid.

[0014] As a further scheme of the present application, in step three, the micro element is further subdivided into triangular facet elements or regular small block elements when implemented in engineering; the scattering field contribution of the facet element is calculated, and the micro element scattering echo is accumulated in the micro element layer to obtain the target scattering echo; wherein the facet integral term is solved by numerical integration or edge integral approximation.

[0015] As a further scheme of the present application, the clustering angle threshold in step four is set by distance: 1.0° for distance 0-3m, 0.5° for distance 3-10m, and 0.25° for distance 10-20m; and the nearest micro element to the radar is retained as a visible micro element in the same direction cluster, and the remaining micro elements are determined as occluded micro elements and output the occlusion ratio and occlusion scattering intensity statistics for error evaluation and correction.

[0016] As a further scheme of the present application, the standard body in step five is a 200mm×200mm metal plate with a thickness of 32mm, a frequency band of 77GHz, and a calibration distance of 1m; the near-field compensation coefficient kc is obtained by aligning the reference near-field scattering distribution and the model calculation distribution at multiple azimuth angles and finding the proportion coefficient with the smallest overall deviation, and the kc is used to uniformly compensate the visible micro element echo and complete the target RCS synthesis.

[0017] As a further scheme of the present application, the radar position scanning in step six satisfies: distance p takes 1-20m with a step of 0.25m, azimuth angle takes 0°-360° with a step of 2°, and height slices are z=0m, 0.5m, 1.0m, 1.5m; the output three-dimensional RCS matrix takes body number, material configuration number, distance index, azimuth angle index and height index as matrix index.

[0018] The technical effect and advantages of the present application of a near-field pedestrian RCS equivalent micro element simulation model are as follows:

[0019] (1) The present application is oriented to the near-field working condition of vehicle-mounted millimeter wave radar, adopts near-field physical optical scattering field calculation to calculate the scattering contribution of visible surface elements, and the distance attenuation and phase change of incident spherical waves enter the calculation link with the radar position, which can represent the law of pedestrian scattering characteristics changing with distance, azimuth angle and height in the range of 1m-20m.

[0020] (2) The present application uses equal-area equivalent microelements as unified near-field scattering units, records the center, normal, equivalent size parameters, region identification and material identification in the microelement layer, and the six-region curved surface is discretized with consistent scale, so that the geometric, material and shielding information can be organized in the same unit, facilitating batch scanning of microelement sets and maintaining the repeatability of the results.

[0021] (3) The present application constructs a shielding visibility judgment in the microelement layer, clusters the microelements according to the direction of the radar and retains the nearest microelement in the same direction cluster as the visible microelement; the remaining microelements are included in the shielding set and the shielding RCS statistical quantity is output, which is used for error evaluation and correction of the synthesis result in calibration compensation, the shielding set is not involved in the scattering superposition of visible microelements by default, and the incident intensity is modified by introducing a material factor, so that the influence of human body self-shielding and clothing difference can be explained and reproduced.

[0022] (4) The present application scans the radar position according to the deterministic sampling scheme of distance, azimuth angle and height and outputs the scattering characteristic matrix, the matrix index contains body shape, material configuration, distance, azimuth angle and height information, which can be directly used for virtual testing, algorithm evaluation and data comparison, reducing the repeated measurement workload and improving the engineering verification efficiency. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 The present application model overall flow and data matrix generation flow diagram;

[0024] Figure 2 The present application is a schematic diagram of two types of pedestrian body shape;

[0025] Figure 3 The present application is a schematic diagram of shadow shielding of pedestrian target parameterized modeling;

[0026] Figure 4 The present application is a schematic diagram of spatial data matrix and tangent plane scattering intensity. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0028] Example 1

[0029] This embodiment gives an implementation of a near-field pedestrian RCS equivalent micro-element simulation model under the working condition of 77GHz vehicle-mounted millimeter wave radar. The pedestrian is fixed near the origin of the spatial rectangular coordinate system O-XYZ, the Z-axis is vertically upward, and the XY plane is the ground. As shown in Figure 1 , the model is executed according to the link of "geometric parameterization modeling, equivalent micro-element generation (this embodiment preferably circular flat micro-element), near-field scattering calculation, material factor correction, occlusion visibility screening, near-field compensation, data matrix output".

[0030] The unified spatial rectangular coordinate system O-XYZ is used in the application process of the model, the Z-axis is the vertical direction, the XY plane is the ground plane, the pedestrian model is fixed near the coordinate origin, the radar transmitting point position changes with distance and angle, the frequency is 77GHz, the electromagnetic wave propagation speed in free space is 3.00×10 8 m / s, the working wavelength is determined by the frequency and the propagation speed, the wave number is given by , the wave number is always represented by k; the correction coefficient introduced in the synthesis of scattering characteristics is represented by k c . The model output scattering characteristics is denoted as σ, the linear quantity unit is m 2 , which needs to be converted to dB quantity when displayed. Linear quantities are accumulated and corrected during the calculation process, and logarithmic conversion is performed in the final output stage to ensure that the linear superposition relationship in the synthesis process is not changed by the logarithmic operation.

[0031] The pedestrian shape is constructed by parameterized geometric body combination, the human body surface is divided into six regions of head, torso, left arm, right arm, left leg and right leg, the head is described by a sphere, and the torso and limbs are described by cylinders. To reflect the influence of body type difference on near-field RCS and ensure reproducibility, this embodiment introduces two types of body types as input samples, pedestrian No. 1 is "tall and fat", and pedestrian No. 2 is "short and thin", the body type is shown in Figure 2The parameters of the No. 1 pedestrian body type (high and fat) are as follows: the head sphere radius is 0.15 m; the trunk cylinder bottom surface radius is 0.20 m, and the height is 0.60 m; the left arm and the right arm cylinder bottom surface radius is 0.10 m, and the height is 0.80 m; the left leg and the right leg cylinder bottom surface radius is 0.10 m, and the height is 0.90 m. The parameters of the No. 2 pedestrian body type (short and thin) are as follows: the head sphere radius is 0.13 m; the trunk cylinder bottom surface radius is 0.20 m, and the height is 0.50 m; the left arm and the right arm cylinder bottom surface radius is 0.08 m, and the height is 0.70 m; the left leg and the right leg cylinder bottom surface radius is 0.10 m, and the height is 0.80 m. Both groups of body types adopt consistent spatial arrangement rules: the trunk cylinder bottom surface center is located at (0, 0, 0), and the trunk axis is along the positive direction of the Z axis; the head sphere center is located above the trunk and coaxial with the trunk axis; the arm cylinder axis is arranged along the Z direction and connected to the upper left and right sides of the trunk; and the leg cylinder axis is arranged along the Z direction and connected to the lower left and right sides of the trunk.

[0032] In order to directly realize the geometric position, the embodiment gives a group of connection reference points for the No. 1 pedestrian: the head sphere center (0, 0, 0.75), the left arm bottom surface center (0.30, 0, 0.60), the right arm bottom surface center (-0.30, 0, 0.60), the left leg bottom surface center (0.10, 0, -0.90), and the right leg bottom surface center (-0.10, 0, -0.90); the No. 2 pedestrian can be set according to the same structure, the head sphere center is (0, 0, 0.63), the left leg bottom surface center is (0.12, 0, -0.80), the right leg bottom surface center is (-0.12, 0, -0.80), and the rest of the connection points are consistent with the trunk height ratio. The above parameterized configuration ensures that different body type models have comparability in the same coordinate system, and facilitates outputting a data matrix with consistent index rules in the same radar position sampling set.

[0033] The six-region curved surface is discretized into a set of scattering units composed of equivalent microelements. In the embodiment, the equivalent microelements are preferably circular plate microelements. The microelement set satisfies the same area constraint in each region. Each equivalent microelement (preferably a circular plate microelement) in the same region has consistent area. Different regions can use different equal-area targets to adapt to the curvature and size differences. Each microelement at least records: the microelement center position vector P 微元 , the microelement normal unit vector, the microelement equivalent radius r, the region identifier, the material identifier, and the occlusion state.

[0034] When using equal-area equivalent microelements (preferably circular plate microelements), the microelement equivalent radius is determined by the microelement area conversion. The microelement equivalent radius is Where A is the area-equivalent target of the region. The center of the microelement is the centroid of the facet of the region surface after the surface is divided, and the normal of the microelement is the outward normal of the surface at the point. The spherical region is divided into area-equivalent rings: the spherical surface is divided into several rings according to the polar angle, and each ring has an equal area; each ring is further divided into several segments according to the circumferential direction, so that each facet has an area close to the area-equivalent target; and each facet corresponds to a circular microelement in this embodiment.

[0035] To control the calculation complexity and ensure the speed of near-field simulation, the area-equivalent target and the number of microelements of the pedestrian No. 1 are configured as follows: the area-equivalent target of the head region is about 3.0 x 10 -4 m 2 / facet, and the number of microelements is 900-1100; the area-equivalent target of the torso region is about 4.0 x 10 -4 m 2 / facet, and the number of microelements is 1800-2200; the area-equivalent target of the single arm region is about 3.5 x 10 -4 m 2 / facet, and the number of microelements is 700-900; and the area-equivalent target of the single leg region is about 4.5 x 10 -4 m 2 / facet, and the number of microelements is 900-1100. The pedestrian No. 2 is configured with the area-equivalent target and the number of microelements according to the same method, the number of microelements of the head is slightly reduced, the number of microelements of the arm is slightly reduced, and the number of microelements of the torso and the leg is adjusted according to the proportion of the surface area. The area-equivalent discrete manner in the region ensures that each microelement has a uniform geometric scale, which facilitates the use of a unified near-field scattering calculation formula and the use of consistent angular resolution in the occlusion screening.

[0036] After the above microelement set is established, the model calculates the near-field scattering contribution of the visible microelements at each radar sampling position and synthesizes the overall scattering characteristics of the pedestrian. The near-field scattering field is calculated in the form of physical optics near-field integration, and the scattering field expression is as follows:

[0037]

[0038] Where k0 is the wave number; η0 is the wave impedance in free space; J s is the surface induced current, M s is the surface magnetic current; n is the facet number; lv is the number of visible facets participating in the calculation; ρ n is the distance from the radar transmitting point to the center of the nth facet; is the unit direction from the facet to the radar; is the unit incident vector of the incident wave at the facet. Sn Let r be the area of ​​the nth face element; ′ It represents a local position variable within a surface element; This is the near-field incident phase term.

[0039] During engineering operation, each equivalent micro-element (preferably a circular flat plate micro-element) in the micro-element set is composed of several equivalent small surface elements. These equivalent surface elements can be divided using triangular or regular small surface elements. The scattering field contribution of each equivalent surface element is calculated according to the aforementioned scattering field expression. These contributions are then accumulated across the micro-element layer to obtain the micro-element scattering field. Finally, these are superimposed on the visible micro-element set to obtain the overall scattering field of the pedestrian. This process causes the near-field spherical wave phase and distance attenuation to be reduced by ρ. n , In the calculation, the amplitude and phase of the scattered field change with the radar position, thus reflecting the range sensitivity of the echo intensity under near-field conditions.

[0040] The integral term of the surface element is solved numerically using numerical integration or marginal integration approximation. In this embodiment, the Gardon integral approximation can be used:

[0041]

[0042] in, r is the unit vector normal to the surface element. m Let a be the position vector of the midpoint of the m-th edge of the equivalent triangular element; m ω represents the length and direction of the corresponding side. n w n Here are the vector parameters related to the incident and scattering directions. This approximation transforms the exponential integral over the surface element into a finite-term computable expression, making the scattering contribution of the surface element numerically stable. By repeating the above calculation for the equivalent surface element within each infinitesimal element and summing the contributions within the infinitesimal element, the scattering field of that infinitesimal element can be obtained.

[0043] When using an equivalent micro-element with equal area (preferably a circular flat plate micro-element), the relationship between its backscattering RCS and the incident angle is expressed in a closed form to ensure a clear relationship between the micro-element size, the incident angle, and the scattering amount:

[0044]

[0045] in, r is the equivalent radius of the infinitesimal element (equivalent to its radius in the implementation of a circular flat plate infinitesimal element); θ is the angle between the incident direction and the normal of the infinitesimal element; J1(·) is the first-order Bessel function. In numerical implementation, β = 2krsinθ is set. When β is close to 0, the limiting form is used to ensure computational stability; when β is large, the standard Bessel function is used. The infinitesimal element backscattering expression and the near-field PO scattering field calculation are used together to calculate the scattering contribution of the infinitesimal layer and to constrain consistency, ensuring that the infinitesimal scattering presents the correct trend with changing viewing angle. Pedestrians wear clothing in real road environments, and the material properties of different areas of the human body are different. The material differences are adjusted by the material factor to incorporate the incident wave intensity into the scattering calculation. The material factor is expressed by the following formula:

[0046]

[0047] Where ξ is the dielectric constant of the material, μ is the magnetic permeability of the material, and j is the imaginary unit. The material factor affects the incident wave intensity, and the incident wave intensity correction is achieved using:

[0048] E i '=R T ×E i

[0049] Among them, E i E represents the unadjusted incident wave intensity. i ’ To account for the influence of materials on the incident wave intensity, material parameters are configured in tabular form. The dielectric constants of cotton, wool (silk), polyester, and skin are set according to Table 1, and the magnetic permeability is approximated as μ0 for non-ferromagnetic materials. Fixed rules are used for region material mapping: skin is the default for the head; cotton for the torso; and polyester for the limbs. When simulating different clothing combinations, the geometry remains unchanged; only the material identifiers for the corresponding regions are replaced, and the ξ value in the material factor calculation is updated. The incident wave intensity E after material factor correction is shown below. i The current and excitation terms in the near-field PO scattering field calculation are incorporated to bring the influence of material differences on the scattering contribution into the calculation process from the source.

[0050] Table 1 Electromagnetic Parameter Configuration Table for Common Clothing Materials

[0051]

[0052]

[0053] Human self-occlusion has a significant impact on echo contribution in the near field; the model performs occlusion screening at the micro-element level. For each equivalent micro-element (circular in this embodiment), its pointing vector to the radar is calculated using:

[0054] V = P 微元 -P 雷达

[0055] Among them, P 微元 Let P be the center position vector of the infinitesimal element. 雷达 Let V be the radar's emission point position vector. Normalize V to obtain a unit direction vector. Group these unit direction vectors, grouping micro-elements with direction angles less than a threshold α into the same direction cluster. Within each direction cluster, calculate the distance from the center of each micro-element to the radar. Retain the micro-element with the smallest distance as the visible micro-element, and classify the remaining micro-elements as occluded micro-elements and group them into the shadow region. Consider near-field self-occlusion, such as... Figure 3 As shown, for each micro-element, calculate its direction unit vector pointing to the radar, and filter the visible set according to the rule that "if the direction angle is less than the threshold α, they are clustered into the same cluster, and only the micro-element closest to the radar is retained as a visible micro-element in the same cluster"; the rest are included in the occlusion shadow excess item.

[0056] The threshold α is fixed in segments based on distance: 1.0° for distances of 0–3m, 0.5° for distances of 3–10m, and 0.25° for distances of 10–20m. These thresholds are matched to the micro-element size configuration, ensuring that the division of directional clusters is consistent with the expansion of micro-element angles. This suppresses occlusion caused by excessive segmentation of directional clusters in the near-field segment and excessive occlusion caused by excessive merging of directional clusters in the far-field segment. After screening, a visible micro-element set and an occluded micro-element set are obtained: the visible micro-element set participates in the superposition of scattered fields; the occluded micro-element set outputs the occlusion RCS statistics, used for error assessment and correction of the synthesis results during calibration compensation.

[0057] After obtaining the visible micro-scattering contributions from each region, the model synthesizes the overall electromagnetic scattering characteristics of the pedestrian. The scattering characteristics of the six regions—head, left arm, right arm, torso, left leg, and right leg—are denoted as σ. 头部 σ 左臂 σ 右臂 σ 躯干 σ 左腿 σ 右腿 The scattering characteristics of the blocked portion are denoted as σ. 遮挡部分 The overall composite scattering characteristics of pedestrians are as follows:

[0058] σ 合成 =k c (σ 头部 +σ 左臂 +σ 右臂 +σ 躯干 +σ 左腿 +σ 右腿 -σ 遮挡部分 )

[0059] Where, k cThe correction coefficient is used to comprehensively describe the influence of the skin effect and model error on the synthesis results. Near-field and far-field transformation employs NF2FF extrapolation: using the complex electric field / equivalent surface current calculated from the near-field scattering of the metal plate as input, a spherical observation surface is established at a distance R = 1m from the target, and the near-field response is obtained by sampling at the azimuth angle; subsequently, the equivalent current radiation integral is used to extrapolate the near-field to the far-field (or equivalently, to convert the 1m spherical field to the far-field RCS), and the results at 1m are aligned and compared with the theoretical far-field RCS on the same angular grid to obtain k. c .

[0060] The correction factor was obtained through calibration using a standard metal plate and used under the same frequency band and distance conditions. The standard was a 200mm × 200mm metal plate, 32mm thick, operating at 77GHz, at a distance of 1m. Reference near-field scattering distributions and model-calculated distributions were obtained for the standard at multiple azimuth angles. The two were aligned in the angular dimension, and the proportionality coefficient with the minimum overall deviation was calculated as k. c The calibration value. After calibration, k c Used for calculating the synthetic scattering characteristics of pedestrians, ensuring consistency in the synthetic results in terms of magnitude. If further differentiation in surface electromagnetic response at different locations is needed, k can be used. c Expanded to a part-weighted form and obtained using the same calibration approach, a single global k is used by default. c This can satisfy the requirements for generating near-field matrices.

[0061] The model scans the radar position and generates a spatial scattering characteristic matrix. The pedestrian is fixed at the origin. The radar samples in the horizontal plane according to its polar radius and azimuth, and samples at different heights. In this embodiment, the radar position in the horizontal plane is given by the following formula: the radar is located at... in ρ∈[0, 20]m. The preferred height slices are z = 0m, 0.5m, 1.0m, and 1.5m to form a three-dimensional sample; this embodiment preferably shows the polar coordinate distribution of the two slices z = 0m and z = 1.0m. The sampling step size in this embodiment is set as follows: ρ = 1~20m, step size 0.25m; azimuth angle is 0°~360°, step size 2°; height is z = 0 and 1.0m to form a three-dimensional sample, such as... Figure 4 As shown in the diagram, the middle diagram illustrates the relationship between the coordinate system and the sampling plane, where the x-axis represents the forward direction of the vehicle, the y-axis represents the lateral direction, and the z-axis represents the height direction; the blue area represents sampling in the z=0 plane, the red area represents sampling in the z=1 plane, and the arrows indicate that the radar performs near-field scanning in different azimuth angles φ. Figure 4The four polar coordinate subplots correspond to: Model 1 in the z=1 plane, Model 1 in the z=0 plane, Model 2 in the z=1 plane, and Model 2 in the z=0 plane. The different colored curves in each subplot represent the echo intensity distribution at different plane distances from the sampling point (e.g., 5m, 10m, 20m, etc., covering the range of 5m to 20m).

[0062] For each sampling point, the model calculates the infinitesimal vector, incident angle, and range phase based on the radar position, performs material factor correction, occlusion filtering, and near-field scattering calculations, and obtains σ. 合成 The data is then written into a matrix. The matrix index contains the body type number (pedestrian 1 or pedestrian 2), material configuration number, distance index, azimuth index, and height index. The matrix elements store the linear quantity of the synthetic scattering characteristics and the corresponding dB quantity. Through this matrix, the echo intensity distribution at different locations in the near-field space can be directly obtained. Height slices at Z=0 and Z=1 can be used to observe the changing trend of the pedestrian scattering center on different height planes.

[0063] When running the model under the aforementioned real-world conditions, the larger pedestrian (Pedestrian 1) exhibits higher overall composite scattering characteristics than Pedestrian 2 at the same distance and azimuth, with the difference varying with distance and azimuth. At close range (e.g., 2m–3m), the scattering characteristics are more sensitive to distance changes due to the rapid changes in the spherical wave phase and the relationship with obstruction. These changes gradually level off as the distance increases. Frontal scattering is generally stronger at different azimuth angles, while side scattering is relatively weaker, and obstruction filtering can cause significant weakening in certain directions behind the vehicle. When the material configuration changes, the material factor is entered into the scattering calculation link through incident intensity correction, leading to interpretable differences in scattering characteristics under the same geometric conditions. By generating multiple sets of matrices for different body sizes and material configurations, a near-field scattering characteristic database for virtual testing of vehicle-mounted millimeter-wave radar can be formed, enabling the characterization and rapid reproduction of scattering patterns of pedestrian targets within the close-range range.

[0064] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0065] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A near-field pedestrian RCS equivalent infinitesimal simulation model, characterized in that, The simulation model takes a same-area equivalent micro-element set as a unified near-field scattering calculation unit and a unified data carrier, limits pedestrian geometry, material, shielding and compensation to the same micro-element layer to complete closed-loop solving, and avoids parameter inconsistency caused by module concatenation; each micro-element in the micro-element set solidifies static information; Radar connection direction vectors and distances are calculated in real time from radar positions and micro-element center coordinates, and a reproducible three-dimensional RCS data matrix is directly output therefrom; the model is generated and output by the following steps: step one, establishing a pedestrian parameterized geometry divides a head, a torso, upper limbs and lower limbs into curved surface regions; step two, equal-area equivalent micro-elements are respectively used in each curved surface region to discretize each curved surface to generate micro-elements and write them into the parameter field according to the same-area rule; step three, under the given conditions of a working frequency and a radar spatial position, near-field spherical wave physical optics is used to calculate scattering echoes from micro-elements, and reflection correction is performed according to material identification; step four, clustering is performed according to the direction of micro-elements to radars, only the nearest micro-element to the radar is kept as a visible micro-element in each cluster, and the rest of the micro-elements perform shielding processing, and the clustering angle threshold is pre-set according to distance segmentation; step five, near-field compensation coefficients obtained by standard body calibration at the same frequency and distance are used to uniformly compensate the echoes of visible micro-elements and synthesize the target RCS; step six, the radar position is scanned on a distance of 1 meter to 20 meters, an azimuth angle of 0° to 360° and multiple height slice grids, and a three-dimensional RCS matrix is output according to distance, azimuth angle and height indexes.

2. The near-field pedestrian RCS equivalent cell emulation model of claim 1, wherein, The static information of the micro-element at least includes micro-element center coordinates, micro-element outer normal, micro-element area, micro-element equivalent size parameters, region identification, material identification and shielding state; the material identification is used to index corresponding material electromagnetic parameters in a material parameter table.

3. The near-field pedestrian RCS equivalent cell emulation model of claim 1, wherein The pedestrian parameterized geometry of step one is composed of six regions, including a head, a torso, a left arm, a right arm, a left leg and a right leg; The head is described by a sphere, and the torso and limbs are described by cylinders.

4. The near-field pedestrian RCS equivalent cell emulation model of claim 1, wherein, Step two includes adopting a ring division method in the spherical region to make the areas of the rings same, and dividing each ring in the circumferential direction to generate micro-elements satisfying the same-area rule; in the cylindrical side region, height direction division is combined with circumferential division to generate micro-elements satisfying the same-area rule; the micro-element center is taken as the centroid of the corresponding curved surface facet, and the micro-element outer normal is taken as the outer normal of the curved surface at the centroid.

5. The near-field pedestrian RCS equivalent cell emulation model of claim 1, wherein, In step three, the micro-element is further subdivided into triangular facet elements or regular block facet elements in engineering implementation; the scattering field contribution of the facet element is calculated, the micro-element scattering echoes are accumulated in the micro-element layer, and the target scattering echoes are obtained by superimposing the visible micro-element set; The facet element integral term is solved by numerical integration or edge integral approximation.

6. The near-field pedestrian RCS equivalent infinitesimal simulation model of claim 1, wherein, The cluster angle threshold in the fourth step is set by distance segmentation: 1.0° for 0-3 m, 0.5° for 3-10 m, and 0.25° for 10-20 m; and the nearest microelement to the radar is kept as a visible microelement in the same direction cluster, and the remaining microelements are determined as blocked microelements and output the statistics of the blocking proportion and the blocking scattering intensity for error evaluation and correction.

7. The near-field pedestrian RCS equivalent infinitesimal simulation model of claim 1, wherein, The standard body in the step five is a 200mm*200mm metal flat plate with a thickness of 32mm, a frequency band of 77GHz and a calibration distance of 1m; the near-field compensation coefficient k is obtained by aligning the reference near-field scattering distribution with the model calculation distribution at multiple azimuth angles and calculating the proportional coefficient with the minimum overall deviation c , k c for uniformly compensating the visible micro-element echo and completing target RCS synthesis.

8. The near-field pedestrian RCS equivalent infinitesimal simulation model of claim 1, wherein, In the sixth step, the radar position scanning satisfies: the distance p is 1-20 m and the step length is 0.25 m, the azimuth angle is 0°-360° and the step length is 2°, and the height slice is z=0 m, 0.5 m, 1.0 m, 1.5 m; the output three-dimensional RCS matrix takes the body number, the material configuration number, the distance index, the azimuth angle index and the height index as the matrix index.

Images