A method based on local multi-source primary scattering to whole machine scattering characteristics

Abstract

A method based on local multi-source main scattering to whole-vehicle scattering characteristics is proposed. This method divides a complex target into multiple target structural regions, acquires high-resolution scattering data for each local region using a multi-frequency, multi-polarization radar system, and establishes a mapping relationship between local and whole-vehicle scattering behavior by combining imaging processing and main scattering source extraction. Furthermore, a dynamic weighting mechanism is introduced to adaptively adjust the influence weight of each region in the whole-vehicle inversion based on its contribution and measurement confidence, thereby improving the stability and robustness of the inversion results. This invention effectively alleviates the feature loss problem caused by limited viewing angle and structural complexity in traditional measurements and is applicable to multiple application scenarios such as aircraft radar scattering modeling, stealth performance evaluation, and electromagnetic characteristic analysis.

Description

Technical Field

[0001] This invention relates to the field of radar target characteristic analysis, specifically to a method based on local multi-source main scattering to whole-machine scattering characteristics. This method can use local scattering information to approximate the whole-machine RCS, effectively overcoming the inherent problems of limited field of view, high cost and blind zone in traditional all-around RCS measurement, and providing a practical new path for obtaining the scattering characteristics of complex large targets. Background Technology

[0002] Radar Cross Section (RCS) is a crucial parameter for measuring a target's radar reflectivity, and it is of great significance for assessing the stealth performance of complex aircraft targets, target identification, and radar system design. Traditional RCS measurement methods typically rely on omnidirectional scanning and measurement of the entire aircraft target. However, complex aircraft targets have intricate structures and diverse scattering characteristics, leading to limitations in the field of view, measurement blind spots, and high costs associated with traditional methods.

[0003] In recent years, with the development of radar imaging technology, local scattering source identification methods based on two-dimensional imaging have gradually become a research hotspot. These methods image local scattering data, extract scattering hotspot information, and attempt to infer the overall radar cross-section (RCS) from local scattering characteristics. However, existing technologies mostly focus on extracting local scattering features, lacking a systematic method to effectively fuse local information and map it to the overall scattering characteristics of the entire radar system. This results in a certain gap between the reconstructed results and the true RCS curve, making it difficult to meet the needs of accurate analysis of complex targets.

[0004] Therefore, how to fuse multi-source local scattering data, dynamically adjust weights, effectively map local scattering hotspots to the overall scattering characteristics, and further achieve RCS inversion of complex targets has become a technical problem that needs to be solved by existing technologies. Summary of the Invention

[0005] The purpose of this invention is to propose a method based on local multi-source main scattering to whole-machine scattering characteristics, which integrates multi-source radar data and dynamic weight control mechanism. By constructing key technologies such as local region division, main scattering source identification, feature mapping and fusion modeling, it can achieve high-precision reconstruction and intelligent analysis of the whole-machine scattering characteristics of complex targets.

[0006] To achieve this objective, the present invention adopts the following technical solution: A method based on the characteristics of local multi-source main scattering to overall scattering includes the following steps: Local scattering feature extraction step S110: The entire system is divided into several target structure regions. Multi-frequency multi-polarization radar is used to collect scattering data, generate local scattering intensity images, identify scattering hotspots with high amplitude, and extract key parameters to achieve independent modeling and contribution analysis of local scattering characteristics. Imaging mapping and feature fusion step S120: The local main scattering source is mapped to the three-dimensional coordinate system of the whole machine, local and global features are extracted, and the correspondence between local and whole machine scattering behavior is established through feature fusion and matching alignment technology. Multi-source RCS inversion and dynamic weight adjustment step S130: Scattering data from different frequency bands, viewing angles, and polarization conditions are jointly inverted, and the contribution allocation of each scattering source is adaptively optimized through a dynamic weight adjustment mechanism. Based on the fusion results and weight allocation, an approximate expression model of the overall RCS characteristics of the target is constructed, thereby realizing the modeling of the overall scattering behavior of complex targets.

[0007] Optionally, it also includes accuracy assessment and statistical analysis steps S140: Error assessment, weight distribution statistics, and contribution analysis are performed on the inversion results to generate an accuracy assessment report, providing feedback for method optimization.

[0008] Optionally, the local scattering feature extraction step S110 specifically includes: Target structure region partitioning sub-step S111: Based on the aircraft's structural characteristics, material distribution, and scattering properties, the entire aircraft is divided into multiple non-overlapping local regions to ensure that the scattering characteristics within each region are relatively independent and complete. Multi-frequency multi-polarization radar data acquisition sub-step S112: In each local area, the radar system collects the scattered signals of that area under different frequency bands and polarization conditions. The entire system is divided into [number] sections. The local region, the first The scattering signal sampling sequence of a local region is denoted as ,in , indicating the first in the region The amplitude of the scattered signal at each sampling point N Indicates the number of sampling points; Local two-dimensional imaging processing step S113: For local areas, an inverse distance-weighted algorithm is used for two-dimensional imaging, and the pixel points... The scattering intensity is as described in formula (1): (1) in, N Indicates the number of sampling points. For the coordinates in the two-dimensional image The scattering intensity of the pixel, weight Defined as: (2) For pixels With the The Euclidean distance between the measurement points is used to set a threshold after imaging. High-intensity regions are screened and local maxima are detected to identify scattering hotspots: (3) Before selection One hot spot serves as the main scattering source; Main scattering source identification and parameter extraction sub-step S114: For multi-view radar measurement data, an observation signal matrix is ​​constructed, and an independent component analysis algorithm is used to separate the signals, extracting the independent signal components of each main scattering source. The observation model can be expressed as a linear mixture form: (4) in, For the observed signal matrix, This is the mixing coefficient matrix, representing the contribution of each source in different observations. The main scattering source signal matrix, For the noise matrix, The target exists A local principal scattering source, ,in Indicates the first The contribution of the signal vectors of each main scattering source to the overall radar cross section (RCS). Defined as: (5) in, The energy norm of the signal. Indicates the first The contribution ratio of each scattering source after normalization.

[0009] Optionally, the imaging mapping and feature fusion step S120 includes: 3D coordinate mapping sub-step S121: Through projection matrix , to two-dimensional imaging coordinates Spatial points mapped to the whole machine's three-dimensional coordinate system Complete the coordinate transformation; Local and global feature extraction sub-step S122: Based on the scattering intensity calculation results And using the distance weighting mechanism of formula (2), the distance relationship between the pixel and the measurement point is measured; By using the threshold determination through formula (3), hotspot regions that meet the intensity conditions are selected to form a set of local main scattering sources; The RCS contribution rate of each scattering source is calculated according to formula (5). To characterize its energy proportion; This forms multidimensional local features, including spatial distribution, energy intensity, and signal independence. Feature fusion and matching alignment sub-step S123: Formula (7) is used to fuse the local features of each main scattering source into the three-dimensional space, thereby realizing the weighted mapping from the two-dimensional image space to the three-dimensional space; (7) in, Representing a point in three-dimensional space Fusion scattering intensity at the location, The normalized weight function is used to measure the weight of the first... Each main scattering source to a spatial point The extent of the impact For the first point mapped to that point The scattering intensity of each main scattering source.

[0010] Optionally, in the coordinate transformation of the three-dimensional coordinate mapping sub-step S121, a geometric constraint optimization algorithm is introduced to fine-tune or correct the measurement error and local deformation.

[0011] Optionally, the multi-source RCS inversion and dynamic weight adjustment step S130 includes: Multi-source RCS inversion calculation sub-step S131: In three-dimensional space point Fusion scattering intensity at the location By using formula (8) for weighted fusion, a unified expression of multi-source scattering intensity can be achieved. (8) It should be noted that the formula (8) yields... The unweighted fusion result representing the scattering intensity of multiple regions is mainly used for subsequent weight optimization (Equation (9)), accuracy evaluation of step S140, and Figure 6 The diagram shown is a visual representation of the spectrum.

[0012] Dynamic weight optimization sub-step S132: To further improve fusion accuracy and model stability, the weight set was adjusted. Dynamic adjustments are made. This set represents the weight parameters to be optimized, where... , corresponding to the The fusion weights for each measurement source or region. The optimization objective is to minimize the error between the fused scattering distribution and the reference data, while controlling the balance of the weight distribution. This is specifically expressed by formula (9): (9) in, Indicates the fusion scattering results Compared with reference scattering distribution Error measurement function between This is a weight regularization term used to prevent excessive weight concentration. It is a regularization parameter used to balance the weights of the error term and the regularization term; Whole machine RCS feature modeling and expression sub-step S133: The contributions of each scattering source are combined into the overall system response using a weighted superposition method, as shown in formula (10). (10) in This represents the fusion scattering intensity at a spatial point in the region. The fusion weight for this region is derived from the aforementioned contribution rate calculation and satisfies the normalization constraint. It is the frequency response function. This is a directional function, representing the contribution of the scattering source at different incident angles.

[0013] Optionally, in the accuracy assessment and statistical analysis step S140, The error assessment is as follows: compare the inversion results with the reference data, analyze the deviation in different regions and the whole, and judge the accuracy of the inversion model; The weight distribution statistics are as follows: statistically analyze the weight allocation of each local scattering source during the inversion process, check for excessive concentration or abnormal deviation, and ensure that the weight distribution is reasonable; The contribution statistics are as follows: assess the actual contribution ratio of each scattering source to the overall scattering characteristics of the machine, and form a contribution ranking. The accuracy assessment report is a compilation of error assessment, weight distribution statistics, and contribution statistics.

[0014] The present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the above-described method based on the local multi-source main scattering to whole-machine scattering characteristics.

[0015] The present invention further discloses a computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the above-described method based on the local multi-source main scattering to whole-machine scattering characteristics.

[0016] In summary, the present invention has the following advantages: 1. Regional measurement overcomes field-of-view limitations: This invention divides the entire machine into multiple local areas for separate measurement and imaging, and then reconstructs the overall features through three-dimensional mapping. This fundamentally avoids the physical blind spots of a single scan, significantly reduces the site and equipment costs for full-size measurement, and provides an economical and feasible new approach for obtaining the scattering characteristics of large targets.

[0017] 2. Dynamic fusion improves inversion accuracy: This invention introduces a dynamic weight adjustment mechanism based on multiple indicators such as signal-to-noise ratio and coverage, which adaptively optimizes the contribution of each local data in the overall inversion, effectively suppresses noise and abnormal interference, and achieves high-precision and robust reconstruction from local features to the overall RCS, solving the problems of insufficient data utilization and poor result consistency of traditional methods.

[0018] 3. Characteristic Model Supporting Engineering Decisions: This invention outputs a parameterized and visualized whole-machine scattering characteristic model, which clearly reveals the spatial distribution, intensity, and contribution of key scattering sources as a function of viewing angle. Further evaluation and statistical analysis can be performed, which can be directly used for locating weak links in stealth design and evaluating performance. This realizes a tool-based upgrade from data inversion to engineering decision support, and its application value is significant. Attached Figure Description

[0019] Figure 1 This is a flowchart of a method based on local multi-source main scattering to whole-machine scattering characteristics according to a specific embodiment of the present invention; Figure 2 This is a flowchart of local scattering feature extraction according to a specific embodiment of the present invention; Figure 3 This is a flowchart of imaging mapping and feature fusion according to a specific embodiment of the present invention; Figure 4 This is a flowchart of multi-source RCS inversion and dynamic weight adjustment according to a specific embodiment of the present invention; Figure 5 This is a partial two-dimensional imaging image of an aircraft according to a specific embodiment of the present invention; Figure 6 These are RCS curves obtained by fusing and inverting two-dimensional images of different regions according to a specific embodiment of the present invention. Detailed Implementation

[0020] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0021] The main point of this invention is that, since the overall scattering behavior of a complex target is determined by the main scattering sources in multiple local regions, the complex target is divided into multiple target structural regions. High-resolution scattering data of each local region is acquired through a multi-frequency, multi-polarization radar system. Combined with imaging processing and main scattering source extraction, a mapping relationship between local and overall scattering behavior is established. Furthermore, a dynamic weighting mechanism is introduced to adaptively adjust the influence weight of each region in the overall inversion based on its contribution and measurement confidence, thereby improving the stability and robustness of the inversion results. This invention effectively alleviates the feature loss problem caused by limited viewing angle and structural complexity in traditional measurements, and is applicable to multiple application scenarios such as aircraft radar scattering modeling, stealth performance evaluation, and electromagnetic characteristic analysis.

[0022] For details, see Figure 1 The flowchart of a method based on local multi-source main scattering to whole-system scattering characteristics according to a specific embodiment of the present invention is shown, including the following steps: Local scattering feature extraction step S110: The entire system is divided into several target structure regions. Multi-frequency multi-polarization radar is used to collect scattering data, generate local scattering intensity images, identify scattering hotspots with high amplitude, and extract key parameters to achieve independent modeling and contribution analysis of local scattering characteristics.

[0023] For details, see Figure 2 Step S110 includes the following sub-steps: Target structure region partitioning sub-step S111: As a complex three-dimensional structure, an aircraft's radar scattering characteristics are a superposition of scattering contributions from multiple different parts. Due to the large size of aircraft and the limited radar field of view, a single measurement cannot cover the entire aircraft, resulting in incomplete scattering data and affecting the accuracy of RCS inversion.

[0024] Therefore, the sub-step is as follows: Based on the aircraft's structural characteristics, material distribution, and scattering properties, the entire aircraft is divided into multiple non-overlapping local regions, typically 10 to 20, to ensure that the scattering characteristics within each region are relatively independent and complete.

[0025] In one specific embodiment The criteria for classification can specifically include: Based on characteristics: Classify based on the aircraft's external features (such as nose, fuselage, wings, tail, cockpit, etc.); Material distribution: Consider the influence of different material types (metals, composites, coatings, etc.) on scattering characteristics in different parts; Scattering characteristics: Based on the known or estimated scattering intensity distribution, high scattering regions (such as edges and corner reflection structures) are separately divided.

[0026] The principles of classification can specifically include: Each region does not overlap with the others and covers the entire surface of the machine; the number of regions is usually 10–20, balancing measurement efficiency and feature independence; ensuring that the scattering mechanism is relatively consistent within each region, which facilitates subsequent modeling.

[0027] Multi-frequency multi-polarization radar data acquisition sub-step S112: In each local area, the radar system collects the scattered signals of that area under different frequency bands and polarization conditions. The entire system is divided into [number] sections. The local region, the first The scattering signal sampling sequence of a local region is denoted as ,in , indicating the first in the region The amplitude of the scattered signal at each sampling point N The number of sampling points is indicated by the Nyquist sampling theorem to avoid signal aliasing. The sampling frequency and the number of sampling points are determined based on the bandwidth of the radar system and the target distance.

[0028] In a specific example, the radar system can have multiple frequency bands, such as S, C, X, and Ku bands, and multiple polarization capabilities, such as HH, HV, VV, and VH. For each predefined local area, the radar beam direction is adjusted for independent scanning.

[0029] Local two-dimensional imaging processing step S113: Since the acquired local scattering signals are composite reflection waveforms of the target at different angles, it is difficult to directly reveal the spatial scattering distribution. Therefore, the inverse distance weighting (IDW) algorithm is used for local two-dimensional imaging. Thus, the steps are as follows: For local areas, an inverse distance-weighted algorithm is used for two-dimensional imaging, and the pixel points... The scattering intensity is as described in formula (1): (1) in, N Indicates the number of sampling points. For the coordinates in the two-dimensional image The scattering intensity of the pixel, weight Defined as: (2) For pixels With the The Euclidean distance between the measurement points is used to set a threshold after imaging. High-intensity regions are screened and local maxima are detected to identify scattering hotspots: (3) Before selection Ten hotspots (typically 10) are used as the main scattering sources. The processed local two-dimensional image is shown below. Figure 5 As shown. In Figure 5 The local two-dimensional image acquisition conditions are: 8-18GHz frequency band, HH polarization, and the parts (a), (b), and (c) are the cockpit, the right wingtip navigation light, and the left wingtip navigation light, respectively.

[0030] Main scattering source identification and parameter extraction sub-step S114: For multi-view radar measurement data, an observation signal matrix is ​​constructed, and the Independent Component Analysis (ICA) algorithm is used to separate the signals and extract the independent signal components of each main scattering source. The observation model can be represented as a linear mixture form: (4) in, For the observed signal matrix, This is the mixing coefficient matrix, representing the contribution of each source in different observations. The main scattering source signal matrix, Let be the noise matrix, which is assumed to be Gaussian white noise.

[0031] The target exists A local principal scattering source, ,in Indicates the first The contribution of the signal vectors of each main scattering source to the overall radar cross section (RCS). Defined as: (5) in, The energy norm of the signal. Indicates the first The normalized contribution ratio of each scattering source indicates that the larger the value, the more significant the influence of that source on the overall scattering characteristics.

[0032] This step enables each local region to output a list of main scattering sources, including the location, intensity, frequency characteristics, and contribution ratio (i.e., normalized contribution rate) of each main scattering source.

[0033] Therefore, this step decomposes the complex whole-system scattering problem into multiple independently analyzable sub-problems, laying the foundation for subsequent whole-system feature reconstruction.

[0034] Imaging mapping and feature fusion step S120: The local main scattering source is mapped to the three-dimensional coordinate system of the whole machine, local and global features are extracted, and the correspondence between local and whole machine scattering behavior is established through feature fusion and matching alignment technology.

[0035] For details, see Figure 3 The imaging mapping and feature fusion step S120 includes: 3D coordinate mapping sub-step S121: To achieve a precise correspondence between the local main scattering source and the overall 3D model, a projection transformation method is employed, using a projection matrix. , to two-dimensional imaging coordinates Spatial points mapped to the whole machine's three-dimensional coordinate system After completing the coordinate transformation, the mapping relationship is as shown in formula (6). (6) in, Indicates the first Coordinates in a two-dimensional image of a local region For the first The projection matrix corresponding to each local region is obtained by combining the radar viewpoint, range, and target attitude. These are the corresponding three-dimensional spatial coordinates.

[0036] Furthermore, in the coordinate transformation, a geometric constraint optimization algorithm is introduced to fine-tune or correct the measurement error and local deformation, ensuring the accuracy of spatial positioning, so that the location of the mapped scattering source is consistent with the actual aircraft structure in space.

[0037] Local and global feature extraction sub-step S122: This sub-step is used to further extract multi-level features of both local and global aspects after completing the three-dimensional spatial mapping of the local main scattering source.

[0038] Based on the scattering intensity calculation results Furthermore, by utilizing the distance weighting mechanism of formula (2), the distance relationship between the pixel and the measurement point is measured, highlighting the contribution of the nearest scattering source. This two-dimensional weight not only ensures the spatial rationality of local features but also provides a foundation for subsequent weight expansion in three-dimensional space.

[0039] By using the threshold determination through formula (3), hotspot regions that meet the intensity conditions are selected to form a set of local main scattering sources; The RCS contribution rate of each scattering source is calculated according to formula (5). To characterize its energy proportion; This forms multidimensional local features, including spatial distribution, energy intensity, and signal independence, providing a foundation for the fusion of overall features.

[0040] In one specific embodiment, the multidimensional local feature can be expressed in the form of a vector.

[0041] Feature fusion and matching alignment sub-step S123: Formula (7) is used to fuse the local features of each main scattering source into the three-dimensional space, thereby realizing the weighted mapping from the two-dimensional image space to the three-dimensional space; (7) in, Representing a point in three-dimensional space Fusion scattering intensity at the location, The normalized weight function is used to measure the weight of the first... Each main scattering source to a spatial point The extent of the impact For the first point mapped to that point The scattering intensity of each main scattering source, weighting function It inherits the basic idea of ​​two-dimensional distance weighting in formula (2) and extends it to three-dimensional space. Specifically, Replacing the two-dimensional distance with a three-dimensional Euclidean distance allows for further consideration of the contribution rate of the scattering source. Signal-to-noise ratio and historical stability indicators The comprehensive weight is constructed from factors such as [list of factors], and its value is normalized as follows:

[0042] The weighting function can be normalized based on actual measurement data to ensure a reasonable contribution ratio of different scattering sources in spatial fusion. Through the above weighted fusion process, not only is local spatial continuity maintained, but the resolution and stability of feature representation are also enhanced, ultimately forming a three-dimensional scattering feature map of the target, laying the foundation for subsequent identification, classification, and high-precision RCS inversion.

[0043] Therefore, this feature fusion process takes into account both local salience and overall spatial continuity, ultimately forming a three-dimensional scattering feature map of the target, providing high-resolution feature support for subsequent identification and classification, and laying the foundation for subsequent high-precision whole-machine RCS inversion.

[0044] Multi-source RCS inversion and dynamic weight adjustment step S130: Scattering data from different frequency bands, viewing angles, and polarization conditions are jointly inverted, and the contribution allocation of each scattering source is adaptively optimized through a dynamic weight adjustment mechanism. Based on the fusion results and weight allocation, an approximate expression model of the overall RCS characteristics of the target is constructed, thereby realizing the modeling of the overall scattering behavior of complex targets.

[0045] For details, see Figure 4 The multi-source RCS inversion and dynamic weight adjustment step S130 includes the following sub-steps: Multi-source RCS inversion calculation sub-step S131: Because the electromagnetic scattering characteristics of an aircraft are affected by multiple factors such as radar frequency, incident angle, and polarization, a single measurement source can often only capture the local scattering characteristics of the target, making it difficult to comprehensively characterize the electromagnetic response behavior of the entire aircraft. To improve the inversion accuracy and spatial coverage, this invention adopts a multi-source data fusion strategy, integrating scattering data from different frequency bands, different observation angles, and different polarization states to construct a multi-dimensional, multi-angle scattering feature system.

[0046] Since each measurement source corresponds to a set of local main scattering sources, its two-dimensional image coordinates are mapped to three-dimensional space through the projection transformation defined by formula (6), forming a spatial positioning result. The scattering intensity distribution calculated using the aforementioned formula (7) is then used. It can obtain the fused scattering spectrum of each measurement source in three-dimensional space.

[0047] This sub-step will involve three-dimensional space. point Fusion scattering intensity at the location By using formula (8) for weighted fusion, a unified expression of multi-source scattering intensity can be achieved: (8) It should be noted that the formula (8) yields... The unweighted fusion result representing the scattering intensity of multiple regions is mainly used for subsequent weight optimization (Equation (9)), accuracy evaluation of step S140, and Figure 6 The diagram shown is a visual representation of the spectrum.

[0048] Formula (8) realizes a unified expression of multi-source scattering intensity, providing a data foundation and spatial support for subsequent dynamic weight optimization and whole-system RCS modeling.

[0049] Dynamic weight optimization sub-step S132: To further improve fusion accuracy and model stability, and to enhance the accuracy and robustness of the fusion model, this sub-step introduces a constrained optimization mechanism for the weight set. Dynamic adjustments are made. This set represents the weight parameters to be optimized, where... , corresponding to the The fusion weights for each measurement source or region. The optimization objective is to minimize the error between the fused scattering distribution and the reference data, while controlling the balance of the weight distribution. This is specifically expressed by formula (9): (9) in, Indicates the fusion scattering results Compared with reference scattering distribution Error metric function between reference scattering distribution This can be obtained through historical measurements, simulation models, or experimental data. This is a weight regularization term used to prevent excessive weight concentration; it can be obtained using the L2 norm or entropy constraints. It is a regularization parameter used to balance the weights of the error term and the regularization term, and can be obtained through cross-validation or empirical adjustment.

[0050] Whole machine RCS feature modeling and expression sub-step S133: After completing the fusion of multi-source data and weight optimization, an approximate expression model of the RCS characteristics of the whole machine is constructed to achieve unified modeling of the overall scattering behavior of complex targets.

[0051] The contributions of each scattering source are combined into the overall system response using a weighted superposition method, as shown in formula (10). (10) in This represents the fusion scattering intensity at a spatial point in the region. The fusion weight for this region is derived from the aforementioned contribution rate calculation and satisfies the normalization constraint. The frequency response function can be obtained by recording the curve of scattering intensity changing with frequency through multi-frequency radar measurements, and then constructing a function fit. The directional function represents the contribution of the scattering source at different incident angles. It can be obtained by changing the radar incident angle and recording the data of scattering intensity changing with direction to construct a directional map. By constructing a local and global feature system, a unified expression of multi-level features is achieved, providing support for subsequent target recognition, performance evaluation, and simulation verification.

[0052] This model reflects the overall scattering characteristics of the entire target and highlights the contributions of key scattering sources, providing support for stealth performance evaluation and electromagnetic property analysis. Furthermore, the model can reconstruct and visualize the overall scattering behavior of complex targets, specifically as follows: Figure 6 As shown.

[0053] Therefore, the dynamic update of the weight set is achieved through steps S131-S133, and the optimized whole-machine scattering intensity distribution and RCS inversion value are obtained based on this weight.

[0054] The present invention further includes: accuracy assessment and statistical analysis step S140: Error assessment, weight distribution statistics, and contribution analysis are performed on the inversion results to generate an accuracy assessment report, providing feedback for algorithm optimization.

[0055] The error assessment involves comparing the inversion results with reference data (such as simulation results or historical measurement data), analyzing the deviations in different regions and the whole, and judging the accuracy of the inversion model.

[0056] The weight distribution statistics are as follows: statistically analyze the weight allocation of each local scattering source during the inversion process, check for excessive concentration or abnormal deviation, and ensure that the weight distribution is reasonable.

[0057] The contribution statistics are as follows: assess the actual contribution ratio of each scattering source to the overall scattering characteristics of the machine, form a contribution ranking, and highlight the role of key scattering sources in the overall scattering behavior.

[0058] The accuracy assessment report integrates the error assessment, weight distribution statistics, and contribution statistics into an accuracy assessment report, providing feedback for algorithm improvement, model correction, and engineering applications.

[0059] Through the above analysis, this step can not only verify the reliability of the inversion results, but also provide data support for subsequent optimization, ensuring that the method has stability and practical value under complex target and multi-source measurement conditions.

[0060] Furthermore, the present invention also discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the above-described method based on the local multi-source main scattering to whole-machine scattering characteristics.

[0061] The present invention further discloses a computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the above-described method based on the local multi-source main scattering to whole-machine scattering characteristics.

[0062] In summary, the present invention has the following advantages: 1. Regional measurement overcomes field-of-view limitations: This invention divides the entire machine into multiple local areas for separate measurement and imaging, and then reconstructs the overall features through three-dimensional mapping. This fundamentally avoids the physical blind spots of a single scan, significantly reduces the site and equipment costs for full-size measurement, and provides an economical and feasible new approach for obtaining the scattering characteristics of large targets.

[0063] 2. Dynamic fusion improves inversion accuracy: This invention introduces a dynamic weight adjustment mechanism based on multiple indicators such as signal-to-noise ratio and coverage, which adaptively optimizes the contribution of each local data in the overall inversion, effectively suppresses noise and abnormal interference, and achieves high-precision and robust reconstruction from local features to the overall RCS, solving the problems of insufficient data utilization and poor result consistency of traditional methods.

[0064] 3. Characteristic Model Supporting Engineering Decisions: This invention outputs a parameterized and visualized whole-machine scattering characteristic model, which clearly reveals the spatial distribution, intensity, and contribution of key scattering sources as a function of viewing angle. Further evaluation and statistical analysis can be performed, which can be directly used for locating weak links in stealth design and evaluating performance. This realizes a tool-based upgrade from data inversion to engineering decision support, and its application value is significant.

[0065] Obviously, those skilled in the art will understand that the various units or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device, or alternatively, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by the computing device. Alternatively, they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0066] The above description is a further detailed explanation of the present invention in conjunction with specific preferred embodiments. It should not be considered that the specific embodiments of the present invention are limited to this. For those skilled in the art, several simple deductions or substitutions can be made without departing from the concept of the present invention, and all such deductions or substitutions should be considered to fall within the scope of protection of the present invention as defined by the submitted claims.

Claims

1. A method based on the characteristics of local multi-source main scattering to overall scattering, characterized in that, Includes the following steps: Local scattering feature extraction step S110: The entire system is divided into several target structure regions. Multi-frequency multi-polarization radar is used to collect scattering data, generate local scattering intensity images, identify scattering hotspots with high amplitude, and extract key parameters to achieve independent modeling and contribution analysis of local scattering characteristics. Imaging mapping and feature fusion step S120: The local main scattering source is mapped to the three-dimensional coordinate system of the whole machine, local and global features are extracted, and the correspondence between local and whole machine scattering behavior is established through feature fusion and matching alignment technology. Multi-source RCS inversion and dynamic weight adjustment step S130: Scattering data from different frequency bands, viewing angles, and polarization conditions are jointly inverted, and the contribution allocation of each scattering source is adaptively optimized through a dynamic weight adjustment mechanism. Based on the fusion results and weight allocation, an approximate expression model of the overall RCS characteristics of the target is constructed, thereby realizing the modeling of the overall scattering behavior of complex targets.

2. The method according to claim 1, characterized in that: It also includes accuracy assessment and statistical analysis steps S140: Error assessment, weight distribution statistics, and contribution analysis are performed on the inversion results to generate an accuracy assessment report, providing feedback for method optimization.

3. The method according to claim 1 or 2, characterized in that: The local scattering feature extraction step S110 specifically includes: Target structure region partitioning sub-step S111: Based on the aircraft's structural characteristics, material distribution, and scattering properties, the entire aircraft is divided into multiple non-overlapping local regions to ensure that the scattering characteristics within each region are relatively independent and complete. Multi-frequency multi-polarization radar data acquisition sub-step S112: In each local area, the radar system collects the scattered signals of that area under different frequency bands and polarization conditions. The entire system is divided into [number] sections. The local region, the first The scattering signal sampling sequence of a local region is denoted as ,in , indicating the first in the region The amplitude of the scattered signal at each sampling point N Indicates the number of sampling points; Local two-dimensional imaging processing step S113: For local areas, an inverse distance-weighted algorithm is used for two-dimensional imaging, and the pixel points... The scattering intensity is as described in formula (1): (1) in, N Indicates the number of sampling points. For the coordinates in the two-dimensional image The scattering intensity of the pixel, weight Defined as: (2) For pixels With the The Euclidean distance between the measurement points is used to set a threshold after imaging. High-intensity regions are screened and local maxima are detected to identify scattering hotspots: (3) Before selection One hot spot serves as the main scattering source; Main scattering source identification and parameter extraction sub-step S114: For multi-view radar measurement data, an observation signal matrix is ​​constructed, and an independent component analysis algorithm is used to separate the signals, extracting the independent signal components of each main scattering source. The observation model can be expressed as a linear mixture form: (4) in, For the observed signal matrix, This is the mixing coefficient matrix, representing the contribution of each source in different observations. The main scattering source signal matrix, This is the noise matrix; The target exists A local principal scattering source, in Indicates the first The contribution of the signal vectors of each main scattering source to the overall radar cross section (RCS). Defined as: (5) in, The energy norm of the signal. Indicates the first The contribution ratio of each scattering source after normalization.

4. The method according to claim 3, characterized in that: The imaging mapping and feature fusion step S120 includes: 3D coordinate mapping sub-step S121: Through projection matrix , to two-dimensional imaging coordinates Spatial points mapped to the whole machine's three-dimensional coordinate system Complete the coordinate transformation; Local and global feature extraction sub-step S122: Based on the scattering intensity calculation results And using the distance weighting mechanism of formula (2), the distance relationship between the pixel and the measurement point is measured; By using the threshold determination through formula (3), hotspot regions that meet the intensity conditions are selected to form a set of local main scattering sources; The RCS contribution rate of each scattering source is calculated according to formula (5). To characterize its energy proportion; This forms multidimensional local features, including spatial distribution, energy intensity, and signal independence. Feature fusion and matching alignment sub-step S123: Formula (7) is used to fuse the local features of each main scattering source into the three-dimensional space, thereby realizing the weighted mapping from the two-dimensional image space to the three-dimensional space; (7) in, Representing a point in three-dimensional space Fusion scattering intensity at the location, The normalized weight function is used to measure the weight of the first... Each main scattering source to a spatial point The extent of the impact For the first point mapped to that point The scattering intensity of each main scattering source.

5. The method according to claim 4, characterized in that: In the coordinate transformation of the three-dimensional coordinate mapping sub-step S121, a geometric constraint optimization algorithm is introduced to fine-tune or correct the measurement error and local deformation.

6. The method according to claim 4, characterized in that: The multi-source RCS inversion and dynamic weight adjustment step S130 includes: Multi-source RCS inversion calculation sub-step S131: In three-dimensional space point Fusion scattering intensity at the location By using formula (8) for weighted fusion, a unified expression of multi-source scattering intensity can be achieved: (8)； Dynamic weight optimization sub-step S132: To further improve fusion accuracy and model stability, the weight set was adjusted. Dynamic adjustments are made, where the set represents the weight parameters to be optimized, where , corresponding to the The fusion weights for each measurement source or region are optimized to minimize the error between the fused scattering distribution and the reference data, while controlling the balance of the weight distribution. The specific content is expressed by formula (9): (9) in, Indicates the fusion scattering results Compared with reference scattering distribution Error measurement function between This is a weight regularization term used to prevent excessive weight concentration. It is a regularization parameter used to balance the weights of the error term and the regularization term; Whole machine RCS feature modeling and expression sub-step S133: The contributions of each scattering source are combined into the overall system response using a weighted superposition method, as shown in formula (10). (10) in This represents the fusion scattering intensity at a spatial point in the region. The fusion weight for this region is derived from the aforementioned contribution rate calculation and satisfies the normalization constraint. It is the frequency response function. This is a directional function, representing the contribution of the scattering source at different incident angles.

7. The method according to claim 6, characterized in that: In the accuracy assessment and statistical analysis step S140, The error assessment is as follows: compare the inversion results with the reference data, analyze the deviation in different regions and the whole, and judge the accuracy of the inversion model; The weight distribution statistics are as follows: statistically analyze the weight allocation of each local scattering source during the inversion process, check for excessive concentration or abnormal deviation, and ensure that the weight distribution is reasonable; The contribution statistics are as follows: assess the actual contribution ratio of each scattering source to the overall scattering characteristics of the machine, and form a contribution ranking. The accuracy assessment report is a compilation of error assessment, weight distribution statistics, and contribution statistics.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method based on the local multi-source main scattering to whole-machine scattering characteristics as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the method based on the local multi-source main scattering to whole-machine scattering characteristics as described in any one of claims 1-7.

Images