Bistatic synthetic aperture radar track planning method and system based on structure driving

Abstract

The invention provides a bistatic synthetic aperture radar track planning method and system based on structure driving, and relates to the technical field of radar track planning, and the method comprises the steps: building a BiSAR echo model based on a target scattering model; analyzing the imaging characteristics of the target under different configurations, and revealing the constraint relationship between the target structure representation and the configuration; constructing a structure-driven multi-stage trajectory planning model based on the constraint relationship between the target structure representation and the configuration; the trajectory optimization is modeled as a multi-objective optimization problem; through a non-dominated sorting genetic algorithm based on target scattering characteristic driving, a multi-target optimization problem is solved so as to plan a platform track in each stage. According to the method, the limitation of a point target model in an existing trajectory planning method can be broken through, all main lobe observation configurations capable of presenting target structure information are optimized, the retention degree of the target structure information in a BiSAR imaging result is effectively improved, and the interpretability of the image is remarkably enhanced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar trajectory planning, and particularly relates to a structure-driven bi-static synthetic aperture radar trajectory planning method and system. BACKGROUND

[0002] At present, the imaging trajectory planning methods of the bi-static synthetic aperture radar (BiSAR) mainly include: one is the planning method oriented to the image resolution, the core goal of which is to optimize the focusing accuracy of the image; two is the optimization method based on the overall quality indicators such as image entropy; three is the path planning method for multi-platform cooperation, which aims to achieve high-quality imaging using the least platform resources. These methods optimize certain specific performances of imaging to a certain extent. However, the optimization targets of these prior arts are generally limited to traditional resolution or image quality indicators, and the underlying model is usually based on the scattering assumption of ideal point targets.

[0003] There is a fundamental limitation in the prior art, that is, the physical law that the scattering characteristics of extended targets (such as ships, buildings) will change significantly with the observation geometry is ignored. Due to the dependence on the point target model, the existing method cannot represent the directional scattering characteristics of typical scattering elements such as line targets and plate targets with angle changes, which leads to the fact that the planned trajectory cannot ensure the coverage of the main lobe region of the target scattering pattern. The direct consequence is that the key structural information (such as profile, size) of the target in the imaging result is seriously lost, and the real target in the scene is only presented as several discrete strong scattering points in the image, which greatly reduces the interpretability and target recognition value of the image.

[0004] That is, in the prior art, the BiSAR trajectory planning method not only has limitations in the model, but also cannot realize efficient and high-precision target structure information acquisition, which is limited in the BiSAR imaging application for target recognition. SUMMARY

[0005] In order to solve the technical problems in the related art, the present application provides a structure-driven bi-static synthetic aperture radar trajectory planning method and system.

[0006] In order to achieve the above purpose, the technical scheme adopted by the present application includes: According to a first aspect of the present application, there is provided a structure-driven BiSAR trajectory planning method, comprising: step S1: establishing a BiSAR echo model based on a target scattering model; step S2: analyzing the imaging characteristics of a target under different configurations, and revealing the constraint relationship between the target structure representation and the configuration; step S3: constructing a structure-driven multi-stage trajectory planning model based on the constraint relationship between the target structure representation and the configuration; step S4: modeling trajectory optimization as a multi-objective optimization problem; and step S5: solving the multi-objective optimization problem to plan the platform trajectory of each stage by using a non-dominated sorting genetic algorithm based on target scattering characteristics.

[0007] Optionally, the step S1 specifically comprises: Step S1-1: establishing a parameterized echo model based on typical scattering elements, the typical scattering elements including a point target, a line target, a flat target and a dihedral target; wherein, for a point target located at a coordinate , the point target echo model is represented as: , wherein, is a target scattering characteristic term, is a phase term caused by a distance history, is a distance frequency variable, is a slow time variable; , wherein, is an imaginary unit, is the speed of light, represents a distance history term, and are the pitch angles of a transmitter and a receiver, respectively, and are the azimuth angles of the transmitter and the receiver, respectively; Step S1-2: establishing an echo expression of an extended target: , wherein, represents a scattering characteristic term of the extended target, represents an amplitude coefficient, represents a sinc function, is a wave number variable and , and are the length and the height of the target, respectively, and are the azimuth angles of a transmitting station and a receiving station relative to the target in an imaging coordinate system, , , It is the target's own azimuth angle. This represents distance-related envelope terms of different types.

[0008] Optionally, step S2 specifically includes: Step S2-1: Analyzing the relative relationship between the target scattering mode and the synthetic aperture, establishing the structural information preservation characteristics when the synthetic aperture includes the main lobe of the target scattering mode and the point imaging characteristics when the synthetic aperture only includes the side lobes of the target scattering mode; Step S2-2: Distinguishing between main lobe observation and side lobe observation, wherein main lobe observation refers to the main lobe region of the synthetic aperture including the target scattering mode, and side lobe observation refers to the side lobe region of the synthetic aperture including only the target scattering mode; Step S2-3: Establishing the main lobe observation constraint condition for preserving target structural information, the mathematical expression of which is: In the formula, Let k denote the main lobe observation constraint function, and k denote the k-th stage of the target structure planning. The solution represents the trajectory planning for the k-th stage, which is a set of optimization variables used to define the motion trajectories of the transmitter and receiver of the bistatic SAR system within this stage. It is a configuration function used to describe the relationship between the observation geometry of a bistatic SAR system and the target azimuth. This represents the azimuth angle of the target in the k-th stage; Step S2-4: Analyze the BiSAR configuration conditions required to achieve main lobe observation under different target azimuth angles.

[0009] Optionally, step S3 specifically includes: Step S3-1: Perform initial imaging, and obtain the imaging result based on the current synthetic aperture echo using a polar coordinate format algorithm; Step S3-2: Estimate the target structure parameters based on the imaging result to obtain the target location. Structural dimensions L, H and azimuth angle .

[0010] Optionally, step S4 specifically includes: Step S4-1: Establishing a first objective function To maximize the representation of target structural information: In the formula, It is in the target structure representation model and the first Azimuth The associated size weighting factor is expressed as , Indicates the azimuth angle Above, all the first The length of each objective structural component; Step S4-2: Establish the second objective function. To optimize imaging resolution: In the formula, and are the range and azimuth ground resolutions, respectively; step S4-3: establishing a third objective function to optimize the resolution angle: ; in which, is the current resolution cell orientation angle, is the ideal resolution angle and ; step S4-4: establishing a fourth objective function to minimize the platform movement distance: The function is expressed as the total displacement of the transmitting platform and the receiving platform from the position at the previous time to the current planning position, in which, and are the end azimuth angle of the transmitter trajectory in the first stage and the start coordinate of the transmitter trajectory in the second stage, respectively, and are the end azimuth angle of the receiver trajectory in the first stage and the start coordinate of the receiver trajectory in the second stage, respectively, and are the time interval between the discrete trajectory points and the sampling point number, respectively.

[0011] Optionally, the step S5 specifically comprises: step S5-1: using a non-dominated sorting genetic algorithm driven by target scattering characteristics SD-NSGA to solve a constrained multi-objective optimization problem, the model being: in which, is the main lobe observation constraint, is the transmitter azimuth angle constraint, is the receiver azimuth angle constraint, and are the azimuth angle ranges of the transmitter and the receiver, respectively; step S5-2: selecting the optimal solution through a constraint dominance relationship and a crowded distance sorting; step S5-3: fusing the results of each stage through multi-stage planning and imaging to obtain the complete structural information of the target, in which, is the final composite imaging result, is an image processing operator, which functions to extract the structured imaging result obtained by the main lobe observation from the , while filtering out the discrete point-shaped imaging result caused by the side lobe observation, is the first​​​​ A single composite imaging result is formed based on the optimized trajectory to collect data, which only contains partial structural information of the target at a specific observation angle.

[0012] According to a second aspect of the present application, a structure-driven BiSAR trajectory planning system is provided, which is applied to the structure-driven BiSAR trajectory planning method in any of the technical solutions of the first aspect of the present application. The system comprises an initial imaging and parameter estimation unit, a trajectory planning modeling unit, an optimization solving unit and a multi-stage control and data fusion unit. The initial imaging and parameter estimation unit is configured to receive BiSAR echo data, perform initial imaging through a polar format algorithm, and estimate structural parameters of the target based on the imaging result, including target position, size and azimuth angle. The trajectory planning modeling unit is connected with the initial imaging and parameter estimation unit, and is configured to establish a target structure representation model according to the estimated target structural parameters, and construct a multi-target trajectory planning optimization model based on main lobe observation constraints, imaging resolution performance indicators and platform motion constraints. The optimization solving unit is connected with the trajectory planning modeling unit, and is configured to solve the multi-target trajectory planning optimization model by using a non-dominated sorting genetic algorithm driven by target scattering characteristics, and output optimal trajectory parameters of the transmitter and the receiver. The multi-stage control and data fusion unit is connected with the optimization solving unit, and is configured to control the BiSAR platform to perform multiple flight observations according to the optimal trajectory parameters, collect echo data under different configurations, and fuse imaging results obtained in each stage, which contain partial structural information of the target, to generate a final imaging result containing complete structural information of the target.

[0013] Optionally, the trajectory planning modeling unit is specifically configured to establish a target structure representation model, and an expression of the target structure representation model is as follows: wherein to respectively represent the target position, size and azimuth angle estimated through initial imaging, different dominant azimuth angles of the target, represent the total number of dominant azimuth angles of the target, to respectively represent a size weight factor corresponding to each dominant azimuth angle, and the value of the size weight factor is the sum of lengths of all target components at the dominant azimuth angle, to respectively represent the length of the i-th target structure component at the dominant azimuth angle . and construct a multi-target trajectory planning optimization model based on main lobe observation constraints.

[0014] ​According to a third aspect of the present application, there is also provided a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, is capable of implementing the steps of the structure-driven BiSAR trajectory planning method according to any one of the first aspect of the present application.

[0015] According to a fourth aspect of the present application, there is also provided a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, is capable of implementing the steps of the structure-driven BiSAR trajectory planning method according to any one of the first aspect of the present application.

[0016] Advantages: 1. Firstly, the method of the present application can break through the limitations of the traditional point target model and achieve active use of the structural characteristics of extended targets. Specifically, most of the existing related trajectory planning is based on a point target scattering model, which regards the target as an ideal point, and the optimization objective of this model cannot reflect the physical structural characteristics of the extended target. Therefore, the planned trajectory cannot guarantee to capture the key geometric information such as the contour and size of the target, resulting in serious loss of structural information in the imaging result. In the present application, step S1 is to establish a BiSAR echo model based on a target scattering model, which contains a more complex model than the point target. This step lays the foundation for the structure-driven method, and by combining with the subsequent steps, the method of the present application no longer regards the target as a non-discriminatory point, but recognizes that it has scattering characteristics varying with the observation angle, and takes this as the starting point for planning. This is a fundamental change from "ignoring the physical characteristics of the target" to "actively using the physical characteristics of the target".

[0017] Secondly, the method of the present application establishes an explicit association and closed-loop optimization between the target structure and the system configuration. Specifically, the traditional trajectory planning is disconnected from the physical structure of the target. The planning system does not know and cannot judge whether the current trajectory is conducive to presenting the intrinsic structure of the target. In the present application, through steps S2 and S3, the target structure is explicitly introduced as a core variable into the planning model. In this way, the planning algorithm actively considers how to adopt a certain configuration to meet the constraint conditions required for target structure imaging when making decisions. At the same time, the solving process of step S5 is to search in this model embedded with structural constraints, thereby realizing the close coupling of trajectory planning and target structure requirements.

[0018] Thirdly, the method of the present application can realize the paradigm conversion from "single static observation" to "multi-stage active perception". Specifically, most of the traditional methods aim to plan an optimal trajectory for a single flight. For a complex target, the single observation angle is limited, which inevitably leads to incomplete structural information acquisition. However, in step S3 of the present application, a "multi-stage trajectory planning model" is explicitly required, and in step S5, "planning the platform trajectory of each stage" is required. This shows that the method is essentially an iterative and active process, which collects structural information of different sides of the target through multiple stages and different configurations in a planned and step-by-step manner. This multi-stage active perception paradigm is the key to ensuring that the complete structural information of the target can be ultimately obtained, and it is beneficial to solve the inherent defect of limited observation angle of single trajectory.

[0019] Fourthly, the method of the present application provides a systematic optimization framework that takes into account both "structural information" and "other performance". Specifically, in the prior art, the resolution-oriented method ignores the structural information, and if only the structure is considered, the imaging quality or platform efficiency may be sacrificed. However, in the method of the present application, step S4 models the trajectory optimization as a multi-objective optimization problem. This shows that the method of the present application does not single-mindedly and at all costs pursue structural information, but rather, it comprehensively weighs the "structural information representation" and other conventional optimization targets in the same framework. Furthermore, through the intelligent optimization algorithm of step S5, a set of Pareto optimal solutions that balance multiple competing targets is ultimately obtained, so that the planned trajectory is both structure-driven and practically feasible and efficient.

[0020] Overall, the method of the present application establishes an echo model based on the scattering characteristics of the target, reveals and utilizes the constraint relationship between the target structure and the observation configuration, and constructs a structure-driven, multi-stage, multi-objective optimization trajectory planning framework. This framework fundamentally changes the drawbacks of traditional BiSAR trajectory planning that ignores the physical characteristics of the target, realizes the leap from passive imaging to active perception, and thus can systematically and efficiently improve the retention degree of target structural information and the interpretability of the image in the final imaging result.

[0021] 2, Other benefits or advantages of the present application will be described in detail in the specific embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiment description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor. Among them: Figure 1is a structural driving-based BiSAR trajectory planning method provided by an exemplary embodiment of the present application, and a step flowchart diagram thereof is shown in FIG. 1; Figure 2 is a schematic diagram of the problems existing in the prior art, which shows the imaging result of a conventional BiSAR on a ship target, wherein, Figure 2 a is an optical image of the ship and an experimental configuration, Figure 2 b is the corresponding BiSAR imaging result, and it can be seen that the target structure information is seriously lost; Figure 3 is a schematic diagram of the BiSAR imaging result and echo extraction result cited in the background of the present application, which takes a lamp post target as an example to show the imaging result and the echo envelope obtained by two-dimensional matched filtering and echo extraction varying with the observation frequency and angle; Figure 4 is a schematic diagram of the theoretical basis of the present application, which shows the relative relationship between the scattering pattern of a linear target and the synthetic aperture, and the two situations of main lobe observation and side lobe observation; Figure 5 is a schematic diagram of a typical geometric configuration of a BiSAR system provided by an exemplary embodiment of the present application, which shows the spatial geometric relationship and key angle parameters of the transmitter and receiver separation; Figure 6 is a schematic diagram of echo acquisition and imaging result of a linear target, which shows the whole process from the original echo to the final imaging result in the two situations of main lobe observation and side lobe observation; wherein, Figure 6 a is the main lobe echo and imaging result of the scattering pattern of a linear target, Figure 6 b is the side lobe echo and imaging result of the scattering pattern of a linear target; Figure 7 is a schematic diagram of the imaging result of a linear target under different starting positions of the synthetic aperture, and through imaging experiments of six different aperture positions, the importance of main lobe observation for structure information preservation is verified, wherein, Figure 7 a is the imaging result when the synthetic aperture starts at -11° and ends at -1°, Figure 7 b is the imaging result when the synthetic aperture starts at -9° and ends at 1°, Figure 7 c is the imaging result when the synthetic aperture starts at -5° and ends at 5°, Figure 7 d is the imaging result when the synthetic aperture starts at -1° and ends at 9°, Figure 7 e is the imaging result when the synthetic aperture starts at 1° and ends at 11°, Figure 7 f is the imaging result when the synthetic aperture starts at 5° and ends at 15°; Figure 8is the overall framework schematic diagram of the structure-driven BiSAR trajectory planning method provided by an exemplary embodiment of the present application, which shows the multi-stage processing flow of the structure-driven BiSAR trajectory planning; Figures 9 to 11 is the comparison of the imaging results obtained by the multi-stage trajectory planning and the circular track BiSAR imaging results of three different targets (target 1, target 2, target 3) respectively, wherein, Figure 9 a is a schematic diagram of the three-dimensional structure model of target 1, Figure 9 b is the initial configuration imaging result of target 1, Figure 9 c is the first stage imaging result of target 1, Figure 9 d is the second stage imaging result of target 1, Figure 9 e is the fourth stage imaging result of target 1, Figure 9 f is the circular track bistatic SAR imaging result of target 1, Figure 10 a is a schematic diagram of the three-dimensional structure model of target 2, Figure 10 b is the initial configuration imaging result of target 2, Figure 10 c is the first stage imaging result of target 2, Figure 10 d is the second stage imaging result of target 2, Figure 10 e is the fourth stage imaging result of target 2, Figure 10 f is the circular track bistatic SAR imaging result of target 2, Figure 11 a is a schematic diagram of the three-dimensional structure model of target 3, Figure 11 b is the initial configuration imaging result of target 3, Figure 11 c is the first stage imaging result of target 3, Figure 11 d is the second stage imaging result of target 3, Figure 11 e is the fourth stage imaging result of target 3, Figure 11 f is the circular track bistatic SAR imaging result of target 3; Figure 12 is the experimental result corresponding to the application of the method of the present application, which shows the structure information recovery process of a complex ship target through seven-stage trajectory planning, wherein, Figure 12 a is a schematic diagram of the three-dimensional structure model of the ship target, Figure 12 b is the initial configuration imaging result of the ship target, Figure 12 c is the first stage imaging result of the ship target, Figure 12 d is the second stage imaging result of the ship target, Figure 12 e is the third stage imaging result of the ship target, Figure 12 f is the fourth stage imaging result of the ship target, Figure 12 g is the fifth stage imaging result of the ship target, Figure 12 h is the seventh stage imaging result of the ship target. DETAILED DESCRIPTION

[0023] In the prior art, the trajectory planning method of BiSAR generally includes: a planning method for communication-radar integration (this method aims to optimize imaging performance while ensuring communication quality, but does not consider the scattering characteristics of the target), a planning method for high-maneuverable platforms (this method models the trajectory planning as a high-dimensional constraint optimization problem, but the optimization objective is still limited to traditional image quality indicators), a cooperative planning method for UAV clusters (this method uses the least number of platforms to achieve high-quality imaging, but the calculation is complex and is not combined with the physical scattering mechanism of the target).

[0024] For BiSAR imaging, the quality of the imaging result not only depends on the resolution performance determined by the system configuration, but is also significantly affected by the coupling relationship between the target scattering characteristics and the observation geometry. However, the current BiSAR trajectory planning methods not only have their own limitations, but also ignore the inherent scattering characteristics of the target under the BiSAR configuration, resulting in the loss of a large amount of target structure information in the imaging result. This directly manifests as the real target appearing only as several discrete strong scattering points in the image, and the key geometric information such as the contour and size cannot be identified, which is not conducive to BiSAR image interpretation and target identification.

[0025] For example, in the document "M.-A. Lahmeri, W.R. Ghanem, C. Bonfert and R. Schober, 'Robust Trajectory and Resource Optimization for Communication-Assisted UAV SAR Sensing,' in IEEE Open Journal of the Communications Society, vol. 5, pp.3212-3228, 2024", a trajectory optimization method for communication-assisted UAV SAR sensing is proposed, although the cooperation of communication and radar is considered, the optimization objective is still limited to the traditional resolution indicator, ignoring the inherent scattering characteristics of the target under the BiSAR configuration.

[0026] For example, in the document "Z. Sun, H. Ren, H. Sun, G. G. Yen, J. Wu and J. Yang, 'Terminal Trajectory Planning for Synthetic Aperture Radar Imaging Guidance Based on Chronological Iterative Search Framework,' in IEEE Transactions on Cybernetics, vol. 54, no. 5, pp. 3065-3078, May 2024", a chronological iterative search framework is established for trajectory planning, although high-precision imaging is achieved, but due to the fact that the inherent scattering characteristics of the target under the bistatic SAR configuration are not considered, the effective presentation of the target structure information in the imaging result cannot be guaranteed.

[0027] In summary, in the prior art, the BiSAR trajectory planning method not only has limitations on the model, but also cannot realize efficient acquisition of target structure information, and is limited in BiSAR imaging applications for target recognition. Through research, the inventors believe that to effectively improve the target structure information in the BiSAR imaging result, the key lies in two aspects, one is to establish a coupling relationship model between the target scattering characteristics and the imaging configuration, and the other is to design and optimize the trajectory for the "main lobe observation" requirement of the target at different azimuth angles.

[0028] Therefore, the present application provides a novel solution, i.e. the structure-driven bistatic synthetic aperture radar trajectory planning method of the present application, the method of the present application constructs a structure-driven multi-stage trajectory planning framework (Structure-driven Trajectory Planning, SDTP), establishes a target structure representation model based on a parameterized model of a typical scattering primitive, acquires prior information such as target azimuth angle through initial imaging, solves a multi-objective optimization problem by using a target scattering characteristic-driven non-dominated sorting genetic algorithm (SD-NSGA), plans the trajectory of each stage platform, can break through the limitations of the point target model in the existing trajectory planning method, realize the optimization of the main lobe observation configuration of all the target structure information that can be presented, effectively improve the retention degree of the target structure information in the BiSAR imaging result, and significantly enhance the interpretability of the image.

[0029] The technical solutions of the present application will be described in detail below with reference to the accompanying drawings.

[0030] As Figure 1As shown, according to the first aspect of the application, the embodiment provides a structure-driven BiSAR trajectory planning method, comprising: step S1: establishing a BiSAR echo model based on a target scattering model; step S2: analyzing the imaging characteristics of the target under different configurations, and revealing the constraint relationship between the target structure representation and the configuration; step S3: based on the constraint relationship between the target structure representation and the configuration, a structure-driven multi-stage trajectory planning model is constructed; step S4: modeling the trajectory optimization as a multi-objective optimization problem; and step S5: solving the multi-objective optimization problem to plan the platform trajectory of each stage by using a non-dominated sorting genetic algorithm based on the target scattering characteristics.

[0031] Through the above technical solution, first, the method of the application can break through the limitation of the traditional point target model and realize active utilization of the structure characteristics of the extended target. Specifically, most of the existing related trajectory planning is based on a point target scattering model, which regards the target as an ideal point, and the optimization objectives (such as resolution and image entropy) cannot reflect the physical structure characteristics of the extended target (such as a ship or a building). Therefore, the planned trajectory cannot guarantee to capture the key geometric information such as the outline and size of the target, resulting in serious loss of structure information in the imaging result. In the application, step S1 is to establish a BiSAR echo model based on a target scattering model, which contains a more complex model than a point target. This step lays the foundation for the structure-driven method, and by combining with the subsequent steps, the method of the application no longer regards the target as a non-discriminatory point, but recognizes that it has scattering characteristics (i.e. "structure") that vary with the observation angle, and uses this as the starting point for planning. This is a fundamental change from "ignoring the physical characteristics of the target" to "actively utilizing the physical characteristics of the target".

[0032] Second, the method of the application establishes an explicit association and closed-loop optimization between the "target structure" and the "system configuration". Specifically, the traditional trajectory planning (such as optimizing resolution or image entropy) is disconnected from the physical structure of the target. The planning system does not know and cannot judge whether the current trajectory is conducive to presenting the intrinsic structure of the target. In the application, through steps S2 and S3, the target structure is explicitly introduced as a core variable into the planning model. In this way, the planning algorithm actively considers how to use the configuration (trajectory) to meet the constraint conditions required for target structure imaging when making decisions. At the same time, the solving process of step S5 is to search in this model embedded with structure constraints, thereby realizing the close coupling of trajectory planning and target structure requirements.

[0033] Thirdly, the method of the present application can realize the paradigm shift from "single static observation" to "multi-stage active perception". Specifically, most traditional methods aim to plan an optimal trajectory for a single flight. For complex targets, the limited observation angle in a single observation inevitably leads to incomplete structural information acquisition. However, the method of the present application explicitly requires a "multi-stage trajectory planning model" in step S3 and plans the trajectories of each stage platform in step S5. This indicates that the method is essentially an iterative and active process, which collects structural information of different sides of the target in different configurations through multiple stages in a planned and step-by-step manner. This multi-stage active perception paradigm is the key to ensuring that the complete structural information of the target can be ultimately obtained, and it is beneficial to solve the inherent defect of limited observation angle in a single trajectory.

[0034] Fourthly, the method of the present application provides a systematic optimization framework that takes into account both "structural information" and "other performance". Specifically, in the prior art, resolution-oriented methods ignore structural information, and if only structural information is considered, imaging quality or platform efficiency may be sacrificed. However, in the method of the present application, step S4 models the trajectory optimization as a multi-objective optimization problem. This indicates that the method of the present application does not single-mindedly and at all costs pursue structural information, but rather balances the "structural information representation" with other conventional optimization objectives in the same framework. Furthermore, through the intelligent optimization algorithm of step S5, a set of Pareto optimal solutions that balance multiple competing objectives is ultimately obtained, so that the planned trajectory is both structure-driven and practically feasible and efficient.

[0035] Overall, the method of the present application establishes an echo model based on the scattering characteristics of the target, reveals and utilizes the constraint relationship between the target structure and the observation configuration, and constructs a structure-driven, multi-stage, multi-objective optimization trajectory planning framework. This framework fundamentally changes the drawbacks of traditional BiSAR trajectory planning that ignores the physical characteristics of the target, realizes the leap from passive imaging to active perception, and thus can systematically and efficiently improve the retention degree of target structural information and the interpretability of the image in the final imaging result.

[0036] The structure-driven dual-base synthetic aperture radar trajectory planning method of the present application will be described below in conjunction with an exemplary embodiment. It should be noted that this exemplary embodiment uses model simulation data for experimental verification, and all radar echo data is generated by electromagnetic simulation. Specifically, the GPU-accelerated bounce ray-physical optics method is used to ensure that the echo data accurately reflects the complex scattering characteristics of the extended target, and the operation is performed by a 10th generation Intel Xeon Processor (Skylake, IBRS, 64GB RAM) and a NVIDIA V100 (32GB memory).

[0037] Step S01: Model Foundation and Problem Introduction. This invention aims to solve the core problem of severe loss of target structure information in existing BiSAR imaging. This problem is prevalent in practical imaging, for example, in... Figure 2 In the BiSAR imaging results of the ship target shown, compared with the optical image (a), the traditional imaging result (b) shows the ship as only a few discrete scattering points, and its key structural information such as outline and size is completely unidentifiable.

[0038] S02: Analysis of Scattering Characteristics of Extended Targets. The fundamental reason for the loss of the above structural information is that existing methods ignore the complex scattering characteristics of extended targets. For example... Figure 3 The BiSAR imaging and echo extraction results of the lamppost target shown indicate that its echo envelope is not a constant, but a two-dimensional sinc function that varies with the observation frequency and angle. This phenomenon proves that the echo envelope of an extended target is a function of its azimuth angle, i.e., there exists a specific "scattering mode." The "scattering mode" of a linear target and its two typical relationships with the synthetic aperture (main lobe observation and side lobe observation) are as follows: Figure 4 As shown.

[0039] For typical configurations of BiSAR, please refer to [reference needed]. Figure 5 As shown. For those located at coordinates The echo expression for the extended target is expressed as: In the above formula, The echo signal representing the extended target, For target scattering characteristics, The phase term is caused by distance history. These are the elevation angles of the transmitter and receiver, respectively. These are the azimuth angles of the transmitter and receiver, respectively. Indicates the amplitude coefficient. Represents the Singer function. It is a wavenumber variable and , It is distance frequency. and These are the length and height of the target, respectively. and These are the relative azimuth angles. , , It is the target's own azimuth angle. This represents distance-related envelope terms of different types.

[0040] Step S1: Path planning problem modeling. The change of the platform flight path will cause the change of the imaging resolution performance, and the length of the flight path is also a factor that must be considered in the process of continuous flight, therefore, the present application establishes a dual base SAR path planning model according to the scattering characteristics of the target, the imaging resolution performance and the flight path length.

[0041] Step S11: Target scattering characteristic constraint. According to the analysis of the target scattering characteristics, in the dual base SAR configuration, to realize the illumination of the main lobe region of the target scattering pattern, the core lies in the observation angle relationship of the dual platform. Considering the actual azimuth angle of the target in the scene , the configuration condition that needs to be met to realize the main lobe observation can be expressed as: ; within a complete synthetic aperture, the observation angle that strictly satisfies the above equation is unique. However, to realize the effective coverage of the main lobe of the target scattering pattern, the key lies in ensuring that the observation angle function value can cross the zero point, that is, its maximum value and minimum value are of opposite signs. According to this, the following mathematical constraint condition of main lobe observation is established: ; wherein, represents the main lobe observation constraint function, is a configuration function, and, ; in the formula, k represents the kth stage of target structure planning, represents the solution of the kth stage trajectory planning, which is a set of optimization variables used to define the motion trajectory of the transmitter and receiver of the dual base SAR system in this stage. This constraint condition ensures that under the currently planned aperture, the BiSAR system can illuminate the main lobe region of the target scattering pattern, thereby laying the foundation for reserving the target structure information in the subsequent stage.

[0042] In addition, the placement angle of the target itself will change the angle of the main lobe observation of the scattering pattern, that is, for the targets with the same placement angle, the configuration that realizes the main lobe observation of the scattering pattern is the same. For this purpose, the present application establishes a target structure characterization model (SRM). The SRM classifies the targets according to the target placement angle, and assigns weights based on the length of the targets with the same angle, in order to preferentially reserve the maximum target structure information in single path planning. Since only the azimuth dimension can represent the target structure information, the target height is not included in the SRM weight. Assuming that there are G targets with different azimuth angles in the scene, the expression of the SRM is: In the formula, to respectively represent the different target dominant azimuth angles obtained by the initial imaging estimation, the total number of target dominant azimuths, respectively represent the size weight factor corresponding to each dominant azimuth, whose value is the sum of the lengths of all target components at the azimuth, respectively represent the length of the i-th target structure component at the azimuth In view of the difference in the placement angle of different target components, single path planning can only present part of the azimuth angle of the target structure. Maximizing the presentation of target structure information in the imaging result in each planning iteration is the key to improving the efficiency of path planning. Therefore, based on the SRM, the first objective function is introduced to guide the optimization process: . Wherein, is the size weight factor associated with the i-th azimuth in the target structure representation model, expressed as , , represents the (equivalent) length of all i-th target structure components at the azimuth Step S12: imaging resolution performance constraint. Imaging resolution performance is a key evaluation of SAR image quality, and the main indicators of imaging resolution performance of bistatic SAR are resolution size and the angle between two-dimensional resolutions. Due to the transmit-receive separation, the sensitivity of the two-dimensional resolution of the bistatic SAR image to the configuration change is higher than that of the traditional SAR. Therefore, in the process of path planning, the imaging resolution performance must be guaranteed. The range resolution of bistatic SAR is: ; wherein, represents the system bandwidth, represents the speed of light, and respectively represent the line-of-sight direction unit vector of the transmitter and the receiver pointing to the target, represents the projection of the distance resolution on the ground. The azimuth resolution expression based on the gradient theory can be expressed as: ; wherein, represents the wavelength, represents the synthetic aperture time, and respectively represent the angular velocity vector of the transmitter and the receiver relative to the target. Under the far-field assumption, the following relationship can be obtained: , , , ; wherein, represents the position vector of the transmitter platform in three-dimensional space, ​​​​​​This represents the position vector of the receiver platform in three-dimensional space. This represents the vertical height coordinates of the transmitter platform. This represents the x-axis coordinate of the transmitter platform on the ground plane. This represents the y-axis coordinate of the transmitter platform on the ground plane. The relationship between the receiver's downward viewing angle and azimuth angle is consistent with the above formula. In bistatic SAR images, resolution performance depends not only on the resolution size but also on the area of ​​the resolvable cells. To improve image quality, minimizing the resolvable cell area is a desirable strategy. The resolvable cell area is defined as: ;in, This represents the included angle at two-dimensional resolution. In a bistatic SAR configuration... The expression is: , , ; Based on the above analysis, while presenting target structure information, the objective function for achieving optimal resolution in bistatic SAR imaging can be expressed as: , ;in, The angle is for the ideal resolution.

[0043] Step S13: Path distance constraint.

[0044] In multi-stage trajectory planning, the platform's movement distance is a critical factor, directly determining the timeliness of task execution. Assume the effective operating distances of the transmitter and receiver are respectively... and Under the far-field assumption, the target distance history can be rewritten as: , In multi-stage planning In the segment trajectory, the trajectory planning variables can be expressed as follows: ;in, and They represent the first The endpoint of the transmitter trajectory in stage k and the starting point of the transmitter trajectory in stage k. and Indicates that respectively represent the first The endpoint position of the stage receiver trajectory and the starting position of the k-th stage receiver trajectory, and the time interval between discrete trajectory points. , Let be the number of sampling points. Under the far-field assumption, the circular flight trajectories of the transmitter and receiver can be approximated as uniform linear motion, thus simplifying the variables describing the flight trajectory as follows: ; Planning the next trajectory To reduce redundant path length and ensure the overall trajectory is as short as possible, the path length should be minimized compared to the previous trajectory. The distance between the two points. Therefore, a fourth objective function is introduced: ; wherein, represents the end position of the trajectory of the first stage transmitter, represents the end position of the trajectory of the first stage receiver, represents the end position of the trajectory of the second stage transmitter, represents the end position of the trajectory of the second stage receiver.

[0045] Step S2: Multi-stage path planning. Based on the imaging characteristics of extended targets in bistatic SAR imaging results, the target scattering guided bistatic SAR path planning framework is designed to achieve target main lobe illumination and retain all the structural information that the target can present.

[0046] Since the main lobe observation can only reveal the target structure information in the azimuth direction, and different targets may have variable initial angles, this paper adopts a multi-stage trajectory planning strategy to capture all the observable structure information of the target. Each stage includes three steps: initial imaging, path planning modeling and path planning problem solving, and the overall architecture process is shown in Figure 8 .

[0047] Step S21: Initial imaging. The purpose is to obtain the approximate structure information, placement angle and position of the target as prior information for subsequent path planning modeling. The initial imaging process includes two steps: BiSAR imaging (imaging the bistatic SAR echo collected by the current aperture, this chapter uses PFA for bistatic SAR imaging. The imaging result can be used to determine the range of target structure and position parameters under the current observation angle, and lay the foundation for subsequent target structure parameter estimation) and target structure parameter estimation (according to the imaging result of the current aperture, the structure parameters of the target are estimated in the image domain combined with the typical target scattering model and optimization method. The rough parameters of the target obtained by this process include target position , structure size and initial azimuth angle ).

[0048] Step S22: Path planning modeling. The purpose is to establish a bistatic SAR path planning model based on target parameter estimation results, imaging resolution performance and path constraints, so as to lay the foundation for subsequent path planning solution. Path planning modeling contains three steps: structure representation model (based on target parameter estimation results, classified according to target azimuth angle, and according to target size, weight factor is assigned, so as to establish SRM. SRM is the basis for each iteration of trajectory planning, aiming to maximize the presentation of target structure information in single planning), target structure constraint (extract the target azimuth angle with the maximum size weight factor from SRM, and based on the main lobe observation imaging principle, establish the target structure constraint for the azimuth angle, to retain the target structure information in the subsequent imaging results) and path planning modeling (combine target structure constraint, imaging resolution size and resolution angle, model the path constraints of bistatic SAR transmitter and receiver, and establish the bistatic SAR path planning problem as a multi-objective optimization problem. Maximize the presentation of target structure in the imaging results, while ensuring the imaging performance).

[0049] Step S23: Path planning problem solution. The main goal is to adjust the path of the double platform based on the path planning model. However, targets at different azimuth angles cannot be fully displayed in single planning imaging results. Therefore, multi-stage flight imaging is needed, and target parameters need to be updated continuously. Path planning problem solution contains three steps: path planning problem solution (generate initial population information by combining configuration parameters and signal parameters. Then, NSGA-II is used to solve the trajectory planning problem, and the flight path length, target structure and imaging resolution performance are comprehensively optimized), record echo based on new path (based on path planning results, the transmitter and receiver move along their respective paths to obtain target echoes in the new configuration) and main lobe imaging result output (image the new collected echo data, retain the imaging results of target structure information that meet the main lobe observation principle, and filter out the point-like imaging results that lack structure information).

[0050] Step S3: Bistatic SAR path planning. Step S31: echo recording and imaging. When recording bistatic SAR echoes, assume that the two-dimensional coordinates of the scene and are indexed by and , respectively, where and represent the number of samples along the range direction and the azimuth direction, respectively. The echo signal of the kth stage of path planning is: ; represents the ideal echo response of a single scattering center located at the scene coordinate under a given trajectory configuration ; according to the wave number domain mapping relationship, the echo after wave number resampling can be obtained as: ; wherein, denotes a two-dimensional wave number resampling operation. A two-dimensional FFT is performed on the resampled echo to obtain the imaging result of the first stage: ; wherein, denotes a two-dimensional wave number resampling operation. A two-dimensional FFT is performed on the resampled echo to obtain the imaging result of the first stage: ; wherein, and denote two-dimensional fast Fourier transforms.

[0051] Step S32: Target structure parameter estimation. Based on the imaging result, the approximate coordinate range and size of the target can be determined, thereby improving the efficiency of target structure parameter estimation. A complex target can be decomposed into point targets, linear targets, flat plates, and dihedral angle bodies, and the target imaging result can also be decomposed into imaging results of these typical components. The target parameter estimation problem is thus established as the following optimization model: ; wherein, denotes the structure and position parameters of the typical target is a residual image, defined as: ; Because the amplitudes of the discrete points of the same scattering center in the imaging result remain the same. Therefore, in the target structure parameter estimation process, the amplitudes are used for differentiation. First, the structure parameters of the strongest scatterer in the target region are estimated, and the parameters of all remaining targets are estimated in turn according to the amplitude gradient. After each estimation, the imaging result of the estimated target is subtracted from the original image to generate a residual image, and finally the estimation results of all target parameters are obtained.

[0052] Step S33: Target scattering guided non-dominated sorting guided algorithm. After obtaining the target structure parameters, a target scattering guided non-dominated sorting genetic algorithm is used to solve the path planning problem for the constrained multi-objective optimization problem. The specific steps of the algorithm are as follows: Initialization: Set the platform system parameters and algorithm iteration solving parameters (population size, maximum iteration number, crossover and mutation parameters, etc.); generate an initial population of N individuals by random generation . Generate a child population by mutation and crossover operations.

[0053] Constraint processing: when generating the child population, if it exceeds the boundary constraint and , replace the parent population and the child population with the boundary values. Merge the parent population and the child population to form a new population (containing individuals), calculate the objective function value and constraint violation degree of each individual: ; denotes the main lobe observation constraint function; Constraint dominance relation: When satisfies the constraint condition , its constraint violation is , it is a feasible solution; a feasible solution dominates an infeasible solution, a smaller infeasible solution dominates a larger infeasible solution.

[0054] Selection operation: based on the objective function value and the constraint violation, the constraint fast non-dominated sorting is carried out; the individuals are sorted from to to the next generation parent population ; if the number of individuals in is still less than , the remaining individuals are selected from the next non-dominated front according to the crowding distance sorting. The crowding distance calculation is: ; and respectively represent the values of the individuals immediately after and before in the current non-dominated front on the first objective function, and respectively represent the maximum value and the minimum value of the first objective function in the current non-dominated front; Iteration termination: after obtaining the population , continue to generate the offspring population through mutation and crossover operations; this process is repeatedly executed until the iteration stopping criterion (such as exceeding the maximum number of iterations) is met; finally, the feasible solution is output from the population as the path planning result.

[0055] Step S4: multi-stage iteration and information fusion. In each path planning-based flight process, only the target structure information that can be effectively presented in the structure representation model is retained, and the specific process is as follows: After each path planning iteration is completed, the system first updates the structure representation model and derives the new target azimuth angle; this path planning process is carried out through continuous iteration until the complete representation of the target structure is achieved; after each flight is completed, the structure information contained in the current structure representation model is extracted from the obtained imaging result ; it is fused with the previously retained structure information to maximize the presentation of the target structure information in the final composite imaging result : ; wherein is an image processing operator, which functions to extract and retain the structured imaging results observed by the main lobe from , while filtering out the discrete point-shaped imaging results produced by the side lobe observation, is the first one planning stage, based on the single composite imaging result formed by collecting data from the optimized trajectory, which only contains partial structural information of the target at a specific observation angle; after emptying the target structural information in the structural representation model, the bistatic SAR path planning ends.

[0056] S5: Path planning example and performance analysis. The following will combine specific numerical experiments to illustrate the structure-driven bistatic SAR path planning method of the present application.

[0057] Step one, set up the simulation environment and parameters.

[0058] To verify the effectiveness of the method, the electromagnetic simulation scene shown in Table 1 is established, and all echo data is generated by the GPU accelerated SBR-PO method. The initial azimuth angle of the transmitter and receiver is randomly set, and the initial elevation angle is fixed at 30° and 40°, respectively.

[0059] Table 1 Simulation parameters of bistatic SAR path planning Step two, execute multi-stage trajectory planning and imaging.

[0060] First, three typical targets with random sizes are planned in multiple stages. Based on the initial imaging result, the target structure parameters are estimated, and the target structure representation model (SRM) is established; then, the non-dominated sorting genetic algorithm driven by target scattering characteristics (SD-NSGA) is used to iteratively solve the multi-objective optimization problem, plan the trajectory of each stage, and perform observation and imaging in turn.

[0061] Step three, fuse the imaging results and evaluate the performance.

[0062] Fuse the imaging results obtained in each stage, which contain local structural information of the target, to generate the final imaging result. As shown in Table 2, the structure information presentation of the three typical targets by the present application is better than 95%, which is significantly higher than the circular bistatic SAR imaging method (the best is only 97.27%, and the worst is only 46.67%), effectively avoiding the generation of false structural information.

[0063] Table 2 Comparison of target structure presentation performance of different path imaging Step four, verify the applicability of complex targets.

[0064] Further apply the method of the present application to complex ship targets. After seven-stage trajectory planning, the complete ship profile and key structure are successfully reconstructed from the initial discrete point-like imaging result, fully demonstrating the excellent target structure information enhancement capability of the present application in complex scenarios.

[0065] As Figures 9 to 12 shown in the above experimental results, the present application has a significant effect on enhancing the structural information of the target, which is embodied in the following aspects: Effective recovery from discrete points to complete structure: under the initial random trajectory pattern, the imaging results are only discrete points that cannot be identified (a, 10a, 11a, 12a). After processing by the method of the present application, the complete geometric structure of all targets is clearly reconstructed (e, 10e, 11e, 12h), which proves that the present application can effectively recover the essential structural information of the target. Figure 9 Figure 9 Better than the traditional circular track observation mode: compared with the traditional circular track BiSAR imaging results (f, 10f, 11f), the results of the present application not only have a more complete structure, but also effectively avoid the generation of false structures. This proves that the present application can obtain more realistic and reliable target features through structure-driven planning.

[0066] Universal for various targets: the experiment covers three typical artificial targets and one complex ship target. The results show that the present application can systematically reconstruct the structures of targets from simple to highly complex through limited stage planning (4-7 stages), which verifies the strong universality and robustness of the method. Figure 9

[0067] In summary, through the structure-driven dual-base SAR path planning method of the present application, the limitation of target structural information loss in traditional trajectory planning can be broken through. Through the active design of the target "main lobe observation" configuration and multi-stage iterative optimization, the efficient and complete acquisition of the target structural information is realized, which significantly improves the interpretability of the image and the target recognition ability.

[0068] In summary, through the structure-driven dual-base SAR path planning method of the present application, the limitation of target structural information loss in traditional trajectory planning can be broken through. Through the active design of the target "main lobe observation" configuration and multi-stage iterative optimization, the efficient and complete acquisition of the target structural information is realized, which significantly improves the interpretability of the image and the target recognition ability.

[0069] ​​According to the second aspect of the present application, there is also provided a structure-driven BiSAR trajectory planning system, which is applied to the structure-driven BiSAR trajectory planning method of any of the first aspect of the present application, and comprises an initial imaging and parameter estimation unit, a trajectory planning modeling unit, an optimization solving unit and a multi-stage control and data fusion unit. The initial imaging and parameter estimation unit is configured to receive BiSAR echo data, perform initial imaging by using a polar format algorithm, and estimate structural parameters of a target based on the imaging result, including target position, size and azimuth angle. The trajectory planning modeling unit is connected to the initial imaging and parameter estimation unit, and is configured to establish a target structure representation model according to the estimated target structural parameters, and construct a multi-target trajectory planning optimization model based on main lobe observation constraints, imaging resolution performance indicators and platform motion constraints. The optimization solving unit is connected to the trajectory planning modeling unit, and is configured to solve the multi-target trajectory planning optimization model by using a non-dominated sorting genetic algorithm driven by target scattering characteristics, and output optimal trajectory parameters of a transmitter and a receiver. The multi-stage control and data fusion unit is connected to the optimization solving unit, and is configured to control the BiSAR platform to perform multiple flight observations according to the optimal trajectory parameters, collect echo data under different configurations, and fuse imaging results obtained in each stage and containing target local structure information, to generate a final imaging result containing target complete structure information.

[0070] According to the third aspect of the present application, there is also provided a computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor is capable of implementing the steps of the structure-driven BiSAR trajectory planning method of any of the first aspect of the present application when executing the computer program.

[0071] According to the fourth aspect of the present application, there is also provided a computer readable storage medium having a computer program stored thereon, wherein the computer program is executable on a processor and capable of implementing the steps of the structure-driven BiSAR trajectory planning method of any of the first aspect of the present application when executed by the processor.

[0072] The above merely illustrates the specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any change or replacement within the technical scope disclosed by the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A structure-driven trajectory planning method for bistatic synthetic aperture radar, characterized in that, The method comprises the following steps: Step S1: establishing a BiSAR echo model based on a target scattering model; Step S2: analyzing imaging characteristics of the target in different configurations, and revealing a constraint relationship between a target structure representation and the configurations; Step S3: constructing a structure-driven multi-stage trajectory planning model based on the constraint relationship between the target structure representation and the configurations; Step S4: modeling trajectory optimization as a multi-objective optimization problem; Step S5: solving the multi-objective optimization problem to plan platform trajectories of each stage by using a non-dominated sorting genetic algorithm based on target scattering characteristics.

2. The structure-driven based bi-static synthetic aperture radar trajectory planning method according to claim 1, wherein, The step S1 specifically comprises: Step S1-1: Establish a parameterized echo model based on typical scattering elements, including point targets, line targets, flat plate targets, and dihedral targets; wherein the echo model of a point target located at coordinates is expressed as: wherein is a target scatter property term, is a phase term caused by range history, is a range frequency variable, is a slow time variable; wherein is the imaginary unit, is the speed of light, denotes a distance history item, and are the elevation angles of the transmitter and receiver, respectively, and are the azimuth angles of the transmitter and receiver, respectively. Step S1-2: establishing an echo expression of an extended target: wherein denotes a scattering characteristic term of the extended target, denotes an amplitude coefficient, denotes a singularity function, is a wave number variable and , and are the length and height of the target, respectively, and are the azimuth angles of the transmitting and receiving stations relative to the target in the imaging coordinate system, respectively, , , is the azimuth angle of the target itself, denote different types of distance-dependent envelope terms.

3. The structure-driven based bi-static synthetic aperture radar trajectory planning method of claim 2, wherein, The step S2 specifically comprises: Step S2-1: analyzing a relative relationship between a target scattering mode and a synthetic aperture, and establishing a structure information maintaining characteristic when the synthetic aperture contains a main lobe of the target scattering mode and a point imaging characteristic when the synthetic aperture contains only a side lobe of the target scattering mode; Step S2-2: distinguishing main lobe observation and side lobe observation, wherein the main lobe observation refers to that the synthetic aperture contains a main lobe region of the target scattering mode, and the side lobe observation refers to that the synthetic aperture contains only a side lobe region of the target scattering mode; Step S2-3: establishing a main lobe observation constraint condition for target structure information maintaining, and a mathematical expression thereof is: wherein, represents the main lobe observation constraint function, k represents the kth stage of the target structure planning, represents the solution of the kth stage trajectory planning, which is a set of optimization variables used to define the motion trajectory of the transmitter and receiver of the bistatic SAR system in this stage, is a configuration function used to describe the relationship between the observation geometry of the bistatic SAR system and the azimuth angle of the target, represents the azimuth angle of the target in the kth stage. Step S2-4: analyzing BiSAR configuration conditions required for realizing main lobe observation under different target azimuth angles.

4. The structure-driven based bi-static synthetic aperture radar trajectory planning method of claim 1, wherein, The step S3 specifically comprises: Step S3-1: performing initial imaging, and obtaining imaging results by using a polar coordinate format algorithm based on current synthetic aperture echoes; Step S3-2: target structure parameter estimation based on imaging results, obtaining target position , structure size L, H and azimuth angle .

5. The structure-driven based bi-static synthetic aperture radar trajectory planning method of claim 1, wherein, The step S4 specifically comprises: Step S4-1: Establishing a first objective function Maximizing the target structure information representation: In the formula, It is in the target structure representation model and the first Azimuth The associated size weighting factor is expressed as , Indicates the azimuth angle Above, all the first The length of each target structural component; Step S4-2: Establishing a second objective function to optimize the imaging resolution: wherein and are the range-to-ground resolution and the azimuth-to-ground resolution, respectively; Step S4-3: Establishing a third objective function In terms of optimizing resolution angle: In the formula, current resolution cell orientation angle, is the ideal resolution angle and ; Step S4-4: Establishing a fourth objective function To minimize the platform movement distance: The function represents the total displacement of the transmitting platform and the receiving platform from the position at the last time to the current planning position, in which, and respectively represent the end azimuth angle of the trajectory of the transmitter in the first phase and the start coordinate of the trajectory of the transmitter in the second phase, and respectively represent the end azimuth angle of the trajectory of the receiver in the first phase and the start coordinate of the trajectory of the receiver in the second phase, represents the time interval between the discrete trajectory points, represents the number of sampling points.

6. The structure-driven based bi-static synthetic aperture radar trajectory planning method of claim 5, wherein, The step S5 specifically comprises: Step S5-1: solving a multi-objective optimization problem with constraints by using a target scattering characteristic-driven non-dominated sorting genetic algorithm SD-NSGA, and a model thereof is: wherein is a main lobe observation constraint, is a transmitter azimuth constraint, is a receiver azimuth constraint, and denote the transmitter and receiver azimuth ranges, respectively. Step S5-2: selecting an optimal solution by using a constraint dominance relationship and a crowding distance sorting; Step S5-3: Fusing the results of each stage by multi-stage planning and imaging to obtain the complete structure information of the target, wherein, represents the final composite imaging result, is an image processing operator, and its function is to extract the structured imaging result observed by the main lobe from , while filtering out the discrete point-like imaging results generated by the side lobe observation, is a single composite imaging result formed based on the optimized trajectory in the first planning stage, which only contains part of the structure information of the target under a specific observation angle.

7. A structure-driven trajectory planning system for a bistatic synthetic aperture radar, the system comprising: The system is applied to the structure-driven dual-base synthetic aperture radar trajectory planning method in any one of claims 1-6, and the system comprises: An initial imaging and parameter estimation unit, configured to receive BiSAR echo data, perform initial imaging by using a polar coordinate format algorithm, and estimate structure parameters of a target based on the imaging results, including target position, size and azimuth angle; A trajectory planning modeling unit, connected with the initial imaging and parameter estimation unit, configured to establish a target structure representation model according to the estimated target structure parameters, and construct a multi-objective trajectory planning optimization model based on a main lobe observation constraint condition, an imaging resolution performance index and a platform motion constraint; An optimization solving unit, connected with the trajectory planning modeling unit, configured to solve the multi-objective trajectory planning optimization model by using a target scattering characteristic-driven non-dominated sorting genetic algorithm, and output optimal trajectory parameters of a transmitter and a receiver. A multi-stage control and data fusion unit, connected with the optimization solution unit, is configured to control the BiSAR platform to perform multiple flight observations according to the optimal trajectory parameters, collect echo data under different configurations, and fuse imaging results obtained in each stage and containing local structure information of the target to generate a final imaging result containing complete structure information of the target.

8. The structure-driven based bi-static synthetic aperture radar trajectory planning method of claim 1, wherein, The trajectory planning modeling unit is specifically configured to establish a target structure representation model, An expression of the target structure representation model is: In the formula to respectively represent the estimated values of the target component length in the initial imaging different target dominant azimuths, represents the total number of target dominant azimuths, to respectively represent the size weight factors corresponding to each dominant azimuth, and the value is the sum of the lengths of all target components in the azimuth, to respectively represent the length of the first target structure component in the azimuth . and construct a multi-target trajectory planning optimization model based on a main lobe observation constraint condition.

9. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor, when executing the computer program, can implement the steps of the structure-driven BiSAR trajectory planning method in any one of claims 1-6.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by the processor, can implement the steps of the structure-driven BiSAR trajectory planning method in any one of claims 1-6.

Images