Method, device and medium for inverse synthetic aperture imaging of a complex moving ship target

Abstract

This application relates to a method, equipment, and medium for inverse synthetic aperture imaging (ICSI) of complex moving ship targets, and pertains to the field of microwave remote sensing technology. The purpose of this application is to address the significant shortcomings of traditional ICSI methods when processing complex moving ship targets. This application performs a global coarse search on the echo signals of each range cell of each scatterer of the complex moving ship to obtain a set of candidate phase parameters matching the echo components of each scatterer. The phase parameters in the candidate phase parameter set are then clustered. A quasi-Newtonian local optimization method is used to estimate the clustered candidate phase parameters, obtaining the phase parameters and reflection coefficients of each scatterer of the complex moving ship. Based on the phase parameters and reflection coefficients of each scatterer of the complex moving ship, the echo signals of each range cell of each scatterer are reconstructed and subjected to Fourier transform to obtain an ICSI image of the complex moving ship.

Description

Technical Field

[0001] This application belongs to the field of microwave remote sensing technology. Background Technology

[0002] Inverse synthetic aperture radar (InSAR) is a core technology for acquiring high-resolution images of complex moving ship targets. However, in high sea states, ships experience three-dimensional non-uniform motion (roll, pitch, and yaw) due to wave action. This causes the echo signal to be modeled as a multi-component cubic phase signal (CPS) in the slow time dimension—each scatterer corresponds to a CPS component, and its phase parameters are directly related to the ship's rotational angular velocity and angular acceleration. Traditional InSAR imaging methods have significant limitations when dealing with such targets.

[0003] Application content

[0004] This application aims to address the significant shortcomings of traditional inverse synthetic aperture imaging methods when dealing with complex moving ship targets. It provides an inverse synthetic aperture radar imaging method for complex moving ship targets (exhibiting three-dimensional non-uniform motion of roll, pitch, and yaw) under high sea states. This method can achieve global search and matching of motion parameters of complex moving ship targets at a low signal-to-noise ratio, thus achieving precise focusing imaging.

[0005] The first aspect of this application provides an inverse synthetic aperture imaging method for complex moving ship targets, including:

[0006] A global coarse search is performed on the echo signals of each distance unit of each scatterer of a complexly moving ship to obtain a set of candidate phase parameters that match the echo components of each scatterer.

[0007] Cluster the phase parameters in the candidate phase parameter set;

[0008] The quasi-Newtonian local optimization method is used to estimate the candidate phase parameters after clustering, thereby obtaining the phase parameters and reflection coefficients of each scatterer of a complexly moving ship;

[0009] Based on the phase parameters and reflection coefficients of each scatterer of the complex moving vessel, the echo signal of each distance unit of each scatterer of the complex moving vessel is reconstructed and Fourier transformed to obtain the inverse synthetic aperture image of the complex moving vessel.

[0010] In one possible design, the multimodal parrot optimization algorithm is used to perform a global coarse search on the echo signals of each distance cell of each scatterer of a complexly moving ship.

[0011] In one possible design, the use of the multimodal parrot optimization algorithm to perform a global coarse search of the echo signals of each range cell of each scatterer of a complexly moving ship includes:

[0012] A population of candidate parameter particles is randomly initialized, wherein the particles are the phase parameters of each scatterer of a complexly moving ship;

[0013] Multiple randomly selected behavioral strategies are employed to update the particles until the maximum number of iterations is reached, obtaining a set of candidate phase parameters that match the echo components of each scatterer. These randomly selected behavioral strategies include communication behavior and fear behavior.

[0014] In one possible design, the fitness function of the multimodal parrot optimization algorithm is:

[0015] ,

[0016] in, For fitness value, Let be the candidate phase parameters, and have , For all scatterers, the first Each echo signal component , The total number of signal components. The pulse repetition interval is... It is the imaginary unit.

[0017] In one possible design, the first of all the scatterers Each echo signal component The expression is:

[0018] ,

[0019] in, The total number of scatterers. , For the first The reflection coefficient of each scatterer For radar operating wavelength, This represents the distance between the scattering point and the center of rotation along the radar line of sight. , , For the first The third phase parameters of the scatterer.

[0020] In one possible design, the clustering of phase parameters in the candidate phase parameter set includes:

[0021] Each phase parameter in the candidate phase parameter set corresponding to each scatterer is used as an initial cluster;

[0022] Two clusters that meet the merging criteria are merged into a new cluster, so that the candidate phase parameter set is updated, until the candidate phase parameter set can no longer be updated.

[0023] The merging condition is: the minimum distance between clusters is less than a preset ratio of the minimum average distance between the two clusters.

[0024] In one possible design, the estimation of candidate phase parameters after clustering using the quasi-Newtonian local optimization method to obtain the phase parameters and reflection coefficients of each scatterer of a complexly moving ship includes:

[0025] The candidate phase parameter with the highest fitness value among the candidate phase parameters corresponding to each scatterer after clustering is used as the initial parameter for local optimization.

[0026] The initial parameters are iteratively optimized using a quasi-Newton algorithm until the magnitude of the gradient is less than a preset convergence threshold. The phase parameters of each scatterer of the complex moving ship are obtained, and the fitness values ​​of the phase parameters of each scatterer of the complex moving ship are used as reflection coefficients.

[0027] In one possible design, the reconstruction of the echo signal of each range unit of each scatterer of the complexly moving vessel based on the phase parameters and reflection coefficients of each scatterer includes:

[0028] The echo signal of each range unit of each scatterer of the complexly moving ship is reconstructed using the following formula:

[0029] ,

[0030] in, For the reconstructed first The scatterer's first Each echo signal component The total number of scatterers. , For the reconstructed first The reflection coefficient of each scatterer For radar operating wavelength, For the first The first-order phase parameter of each scatterer , The total number of signal components. The pulse repetition interval is... It is the imaginary unit.

[0031] The second aspect of this application provides an inverse synthetic aperture imaging device for complex moving ship targets. The inverse synthetic aperture imaging device for complex moving ship targets includes a processor and a memory. The memory stores at least one instruction, which is loaded and executed by the processor to implement the inverse synthetic aperture imaging method for complex moving ship targets as described above.

[0032] A third aspect of this application provides a computer storage medium storing at least one instruction, which is loaded and executed by a processor to implement the inverse synthetic aperture imaging method for complex moving ship targets as described above.

[0033] Specifically, the Multimodal Parrot Optimization (MMPO) algorithm is used to achieve synchronous estimation of parameters for multi-component cubic phase signals (CPS), solving the problems of multi-component interference, parameter estimation error accumulation, and imaging defocus in complex moving ship echo signals. This application includes the following beneficial effects:

[0034] 1. This application can accurately estimate the third-order phase coefficient of complex moving ship targets;

[0035] 2. This application enables rapid multi-peak search with a simple process;

[0036] 3. Data processing results show that this application has good noise resistance and performs well when the signal-to-noise ratio is low.

[0037] In summary, this application is applicable to high-resolution imaging of complex ship targets in scenarios such as maritime surveillance and coastal defense security. Attached Figure Description

[0038] Figure 1 The flowchart shows a method for inverse synthetic aperture imaging of complex moving ship targets based on the multimodal parrot optimization algorithm.

[0039] Figure 2 The dot matrix image of the target to be imaged;

[0040] Figure 3(a) shows the one-dimensional range imaging result after range compression using the range Doppler algorithm;

[0041] Figure 3(b) shows the one-dimensional distance imaging result after distance compression using the SPWVD algorithm.

[0042] Figure 3(c) shows the one-dimensional distance imaging result after distance compression using the imaging method described in the embodiment;

[0043] Figure 4(a) is a schematic diagram of the imaging results of the range Doppler algorithm;

[0044] Figure 4(b) is a schematic diagram of the imaging results obtained by the SPWVD algorithm;

[0045] Figure 4(c) is a schematic diagram of the imaging results of the imaging method described in the embodiment. Detailed Implementation

[0046] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. It should be noted that, unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other.

[0047] Traditional inverse synthetic aperture imaging methods have significant drawbacks when dealing with complex moving ship targets:

[0048] Range Doppler (RD) algorithm: relies on the assumption of uniform rotation, cannot adapt to the non-uniform three-dimensional motion of ships, and the image is severely defocused;

[0049] Time-frequency distribution methods (such as SPWVD) suffer from strong cross-term interference, cannot distinguish the CPS components of ship multi-scatterers, and it is difficult to balance resolution and noise resistance.

[0050] Parametric methods (such as CIGCPF) require component-by-component extraction and signal stripping iterative processes, which lead to severe error accumulation and low computational efficiency in complex ship multi-scatterer scenarios.

[0051] In view of this, the embodiments of this application provide an inverse synthetic aperture imaging method for complex moving ship targets, in order to solve the above-mentioned problems. The following will be discussed in conjunction with the appendix... Figure 1 The implementation scheme of this application will be described in detail.

[0052] Specific Implementation Method 1: To further introduce the implementation method of this application, Figure 1 A flowchart of an inverse synthetic aperture imaging method for complex moving ship targets based on the multimodal parrot optimization algorithm is provided, including steps one through three. The numbering of these steps does not necessarily restrict their execution order. Each step is described in detail below:

[0053] A method for inverse synthetic aperture imaging of complex moving ship targets based on the multimodal parrot optimization algorithm includes:

[0054] Step 1: Modeling the echo of multi-component cubic phase signal (CPS) from a ship in complex motion.

[0055] First, to accurately describe the ship's motion attitude, a corresponding coordinate system needs to be defined. This embodiment establishes a local coordinate system that rotates synchronously with the ship's motion. Its origin At the ship's center of mass, the x, y, and z axes represent the ship's length, beam, and the direction perpendicular to the deck, respectively. Simultaneously, a global coordinate system fixed to the radar is established. Its u, v, and w axes are aligned with the x, y, and z axes of the local coordinate system at the initial moment, respectively.

[0056] Secondly, the complex rotational motion of the ship is decomposed in this coordinate system. The ship's motion in high sea states is decomposed into three typical three-dimensional rotational motions: roll (around the x-axis), pitch (around the y-axis), and yaw (around the z-axis). These three motions can be precisely mathematically described using their respective rotation matrices, as follows:

[0057] Rotation matrix corresponding to roll motion for:

[0058] ,

[0059] in, This is the dynamic rotation angle corresponding to the roll motion.

[0060] Rotation matrix corresponding to pitching motion for:

[0061] ,

[0062] in, This is the dynamic rotation angle corresponding to the pitching motion.

[0063] Rotation matrix corresponding to yaw motion for:

[0064] ,

[0065] in, This is the dynamic rotation angle corresponding to the yaw motion.

[0066] Based on this, to accurately reflect the dynamic rotation characteristics of ships induced by waves under high sea states, the concept of dynamic rotation angle is introduced, and its functional relationship with time is as follows:

[0067] ,

[0068] In the formula, Indicates the direction of motion decomposition. , , These represent the directions of roll, pitch, and yaw, respectively. express Directional oscillation amplitude, express Directional cycle, express Direction and initial phase.

[0069] Finally, after completing translation compensation and inverse Fourier transform, the echo signals of all scatterers can be modeled as a multi-component cubic phase signal (CPS) model. This model can express the echo of each scatterer as the sum of signal components with different phase parameters. Its mathematical expression is:

[0070] ,

[0071] in, Indicates the signal component number. The total number of scatterers. , For the first The reflection coefficient of each scatterer The operating wavelength of the radar (e.g., 3cm in the X-band). The pulse repetition interval (PRI) is the pulse repetition time. The distance between the scattering point and the center of rotation along the radar line of sight in the model is... , and These are the three-phase parameters that are directly related to the ship's rotational motion, and these parameters are the core targets that need to be accurately estimated in subsequent processing steps.

[0072] Step 2: MMPO multi-component CPS parameter estimation.

[0073] 1. A global coarse search based on the multimodal parrot optimization (PO) algorithm is performed on the echo of each range cell. The main objective is to quickly locate the high-fitness parameter region that matches the echo components of multiple scatterers from the ship within a broad parameter space. To achieve this goal, a fitness function needs to be defined to quantify the degree of matching between candidate parameters and the true signal components. In this embodiment, a multinomial Fourier transform is used to construct the fitness function for any candidate parameter vector. Its fitness value The calculation is as follows:

[0074] .

[0075] This fitness function is obtained by analyzing the echo signal. Phase conjugate compensation and coherent accumulation are performed, and the magnitude of the result directly reflects the candidate parameters. The ability to focus the echo energy of a scatterer; the higher the value, the higher the matching degree.

[0076] During the search process, the PO algorithm initializes a pool containing... A population of candidate parameter particles (i.e., parameter vectors) is established, and the normalized Euclidean distance between particles is calculated to effectively distinguish particle populations corresponding to different scatterers (i.e., different fitness peaks), preventing all particles from prematurely clustering in the strongest scatterer component. To balance global exploration and local resource extraction capabilities, particle updates employ four randomly selected behavioral strategies (such as communication and fear behaviors), which incorporate the globally optimal particle... Group mean position And the distribution of Levy flights, which enhances global search capabilities. After a set maximum number of attempts (e.g.) After iteration, this stage will output a set of candidate particles that covers the high fitness region of CPS parameters of all major scatterers of the ship.

[0077] 2. Sub-classification based on hierarchical clustering is then performed. The purpose of this step is to effectively group the relatively dispersed candidate parameter particles obtained in the previous stage according to their respective scatterers. This process first treats each particle as an independent initial cluster, and then iteratively merges them based on the compactness within each cluster (Intra-cluster average distance) and the separation between clusters (Inter-cluster minimum distance). Specifically, the merging rule is that when any two clusters... and minimum distance between When the proportion is less than a certain percentage of their internal average minimum distance (by the clustering threshold) If the clusters are considered to belong to the same scatterer's parameter distribution region, they should be merged. This merging process is repeated until no new clusters can be merged. The final output of this stage is... A separate parameter subclass, in which Corresponding to the total number of ship scatterers detected, the particles within each subclass are closely distributed around the true CPS parameters of the same scatterer.

[0078] 3. Finally, precise parameter estimation based on quasi-Newton (BFGS) local optimization is performed. The goal is to refine the parameters in each subclass to achieve extremely high estimation accuracy. For each pre-defined cluster subclass, the particle with the highest fitness value is first selected as the initial point for local optimization, ensuring that the starting position of the optimization is very close to the true solution. Subsequently, the quasi-Newton (BFGS) algorithm is used to iteratively optimize the initial parameters. In each iteration, the algorithm estimates the gradient of the current parameter point using the finite difference method. And combined with the dynamically updated approximate inverse Hessian matrix To determine the optimal descent direction This allows for efficient parameter updates. Among these, the inverse Hessian matrix... The update follows the classic BFGS formula:

[0079] ,

[0080] In the formula, and These are the increments of the parameters and the gradient, respectively. and They are respectively and The transpose of .

[0081] This iterative process continues until the magnitude of the gradient is less than a preset convergence threshold, indicating that the optimal parameter solution for that scatterer component has been found. By performing the same refinement process on all P subclasses, a complete set of high-precision parameter estimates for all CPS components of the ship's scatterers is finally obtained. and fitness function .

[0082] Step 3: Signal reconstruction and image generation.

[0083] After accurately acquiring the CPS phase parameters of all scatterers for each range cell, the next step is to reconstruct these signal components. First, based on each set of accurate parameters output from the previous stage... and reflection coefficient The echo within each range cell is reconstructed, and the echo signal is... It can be refactored as:

[0084] ,

[0085] in, for The first element.

[0086] Then, a Fourier transform is performed along the slow time dimension to complete azimuth compression. The final output of the entire imaging process is a clear, focused inverse synthetic aperture image of the ship.

[0087] To comprehensively evaluate the performance of the proposed MMPO algorithm in estimating cubic phase parameters for ISAR imaging of maneuvering ships, this study simultaneously conducted simulation and field experiments. In the simulation experiments, a multi-scattering-point ship model was used (see...). Figure 2The radar system was configured with complex motion parameters such as roll, pitch, and yaw based on the five sea state standards in Table 1. The radar system operated in the X-band with a bandwidth of 400MHz, transmitting 512 pulses. The signal-to-noise ratio after pulse compression was 0dB to simulate a strong noise interference environment. To verify the practical application effect of the algorithm, measured data of maneuvering ships acquired by an X-band shore-based ISAR radar were further processed. In the experiment, the traditional RD algorithm, SPWVD algorithm, CIGCPF algorithm, and the proposed MMPO algorithm were used to image the simulated and measured data, respectively. The results are shown in Figures 3 and 4, where Doppler cell represents Doppler and Range cell represents the imaging area. To quantitatively evaluate the imaging quality and efficiency, the image entropy and computation time of each algorithm were calculated (see Tables 2 and 3 for details). The overall results show that the RD algorithm resulted in severe image defocusing; the SPWVD algorithm, although able to suppress some defocusing, reduced resolution and blurred the image; the CIGCPF algorithm had better focusing effect, but required a huge amount of computation. In comparison, the proposed MMPO algorithm can generate clearly focused and structurally accurate ISAR images in both simulation and real-world scenarios. At the same time, its computational efficiency far exceeds that of the CIGCPF algorithm, verifying its effectiveness and efficiency in performing high-quality ISAR imaging under complex motion and low signal-to-noise ratio conditions.

[0088] Table 1 Simulation Parameters

[0089] ,

[0090] Table 2 Performance comparison of different algorithms in simulation experiments

[0091] ,

[0092] Table 3. Performance comparison of different algorithms in actual experiments.

[0093] .

[0094] Specific Implementation Method Two: The inverse synthetic aperture imaging device for complex moving ship targets described in this embodiment includes a processor and a memory. The memory stores at least one instruction, which is loaded and executed by the processor to implement the inverse synthetic aperture imaging method for complex moving ship targets as described in Specific Implementation Method One.

[0095] Specific Implementation Method 3: A computer storage medium as described in this embodiment stores at least one instruction, which is loaded and executed by a processor to implement the inverse synthetic aperture imaging method for complex moving ship targets as described in Specific Implementation Method 1.

[0096] While specific embodiments of this application have been described herein with reference to them, it should be understood that these embodiments are merely examples of the principles and applications of this application. Therefore, it should be understood that many modifications can be made to the exemplary embodiments, and other arrangements can be designed without departing from the spirit and scope of this application as defined by the appended claims. It should be understood that different dependent claims and features described herein can be combined in ways different from those described in the original claims. It is also understood that features described in conjunction with individual embodiments can be used in other described embodiments.

Claims

1. A method for inverse synthetic aperture imaging of complex moving ship targets, characterized in that, include: A global coarse search is performed on the echo signals of each distance unit of each scatterer of a complexly moving ship to obtain a set of candidate phase parameters that match the echo components of each scatterer. Cluster the phase parameters in the candidate phase parameter set; The quasi-Newtonian local optimization method is used to estimate the candidate phase parameters after clustering, thereby obtaining the phase parameters and reflection coefficients of each scatterer of a complexly moving ship; Based on the phase parameters and reflection coefficients of each scatterer of the complex moving vessel, the echo signal of each distance unit of each scatterer of the complex moving vessel is reconstructed and Fourier transformed to obtain the inverse synthetic aperture image of the complex moving vessel. A global coarse search is performed on the echo signals of each distance cell of each scatterer of a complexly moving ship using the multimodal parrot optimization algorithm; The method of using the multimodal parrot optimization algorithm to perform a global coarse search on the echo signals of each range cell of each scatterer of a complexly moving ship includes: A population of candidate parameter particles is randomly initialized, wherein the particles are the phase parameters of each scatterer of a complexly moving ship; Multiple randomly selected behavior strategies are used to update the particles until the maximum number of iterations is reached, thereby obtaining a set of candidate phase parameters that match the echo components of each scatterer. The fitness function of the multimodal parrot optimization algorithm is: ， in, For fitness value, Let be the candidate phase parameters, and have , For all scatterers, the first Each echo signal component , The total number of signal components. The pulse repetition interval is... The imaginary unit; The clustering of phase parameters in the candidate phase parameter set includes: Each phase parameter in the candidate phase parameter set corresponding to each scatterer is used as an initial cluster; Two clusters that meet the merging criteria are merged into a new cluster, so that the candidate phase parameter set is updated, until the candidate phase parameter set can no longer be updated. The merging condition is: the minimum distance between clusters is less than a preset proportion of the minimum average distance between the two clusters; The method of estimating candidate phase parameters after clustering using the quasi-Newtonian local optimization method to obtain the phase parameters and reflection coefficients of each scatterer of a complexly moving ship includes: The candidate phase parameter with the highest fitness value among the candidate phase parameters corresponding to each scatterer after clustering is used as the initial parameter for local optimization. The initial parameters are iteratively optimized using a quasi-Newton algorithm until the magnitude of the gradient is less than a preset convergence threshold. The phase parameters of each scatterer of the complex moving ship are obtained, and the fitness values ​​of the phase parameters of each scatterer of the complex moving ship are used as reflection coefficients.

2. The inverse synthetic aperture imaging method for complex moving ship targets according to claim 1, characterized in that, The first of all scatterers Each echo signal component The expression is: ， in, The total number of scatterers. , For the first The reflection coefficient of each scatterer For radar operating wavelength, This represents the distance between the scattering point and the center of rotation along the radar line of sight. , , For the first The third phase parameters of the scatterer.

3. The inverse synthetic aperture imaging method for complex moving ship targets according to claim 1, characterized in that, The reconstruction of the echo signal of each range unit of each scatterer of the complexly moving ship based on the phase parameters and reflection coefficients of each scatterer includes: The echo signal of each range unit of each scatterer of the complexly moving ship is reconstructed using the following formula: ， in, For the reconstructed first The scatterer's first Each echo signal component The total number of scatterers. , For the reconstructed first The reflection coefficient of each scatterer For radar operating wavelength, For the first The first-order phase parameter of each scatterer , The total number of signal components. The pulse repetition interval is... It is the imaginary unit.

4. An inverse synthetic aperture imaging device for complex moving ship targets, characterized in that, The inverse synthetic aperture imaging device for complex moving ship targets includes a processor and a memory, wherein the memory stores at least one instruction, which is loaded and executed by the processor to implement the inverse synthetic aperture imaging method for complex moving ship targets as described in any one of claims 1 to 3.

5. A computer storage medium, characterized in that, The computer storage medium stores at least one instruction, which is loaded and executed by a processor to implement the inverse synthetic aperture imaging method for complex moving ship targets as described in any one of claims 1 to 3.

Images