Array ISAR three-dimensional imaging method based on non-ideal trajectory distributed moving platform

Abstract

The invention relates to an array ISAR (Inverse Synthetic Aperture Radar) three-dimensional imaging method based on a non-ideal track distributed moving platform. The method comprises the following steps: acquiring a group of echo signals with array manifold distortion; preprocessing the echo signal to obtain a preprocessed echo signal; performing feature extraction on the preprocessed echo signal to determine a plurality of strong scattering points; a plurality of multi-dimensional error particles are combined, DOA estimation is carried out on the preprocessed echo signals, a plurality of target angle values are determined, and the multi-dimensional error particles are obtained through calculation of an improved PSO algorithm and used for simulating array manifold distortion in the echo signals; and calculating to obtain an array ISAR three-dimensional image by using the plurality of strong scattering points and the plurality of target angle values. According to the method, the key problem of array ISAR imaging under non-ideal motion can be solved through dynamic error modeling and multi-source information fusion, and the method has high precision, strong robustness and engineering practicability.

Description

Technical Field

[0001] The present invention belongs to the technical field of radar imaging, and particularly relates to an array ISAR three-dimensional imaging method based on a distributed moving platform with a non-ideal trajectory. Background Art

[0002] It is of great significance to use high-resolution radar for early warning detection of the sea. Compared with two-dimensional imaging, three-dimensional imaging can obtain more detailed information of the target and play an important role in fields such as target recognition. Most of the existing three-dimensional imaging technologies can only obtain the overall contour information of the target, which actually belongs to 2.5D imaging. To obtain more detailed information and achieve true three-dimensional imaging, an Array Inverse Synthetic Aperture Radar (ArISAR) is required. However, when using ArISAR to collect data, the antenna array will have non-stationary motion due to environmental influence, resulting in array manifold distortion and affecting the quality of three-dimensional imaging.

[0003] Most of the existing three-dimensional imaging technologies are for stationary targets or simple scenarios and cannot be applied to complex moving targets. Moreover, the influence of array manifold distortion on imaging quality is not considered during imaging, so they do not have universality. Summary of the Invention

[0004] To solve the above problems existing in the prior art, the present invention provides an array ISAR three-dimensional imaging method based on a distributed moving platform with a non-ideal trajectory. The technical problems to be solved by the present invention are achieved through the following technical solutions:

[0005] The present invention provides an array ISAR three-dimensional imaging method based on a distributed moving platform with a non-ideal trajectory, including:

[0006] Obtaining a set of echo signals with array manifold distortion; preprocessing the echo signals to obtain preprocessed echo signals; extracting features from the preprocessed echo signals to determine multiple strong scattering points; combining multiple multi-dimensional error particles to perform DOA estimation on the preprocessed echo signals to determine multiple target angle values, where the multiple multi-dimensional error particles are obtained by using an improved PSO algorithm and are used to simulate the array manifold distortion in the echo signals; and calculating an array ISAR three-dimensional image by using the multiple strong scattering points and the multiple target angle values.

[0007] In some embodiments, combining multiple multi-dimensional error particles to perform DOA estimation on the preprocessed echo signal to determine multiple target angle values includes: using the SBL algorithm to process the preprocessed echo signal in combination with the i-th steering vector to obtain the i-th spatial spectrogram, where i is a positive integer and less than or equal to I; wherein, the i-th steering vector is calculated using the i-th group of multi-dimensional error particles; traversing the i-th spatial spectrogram to extract multiple spatial spectral peaks; substituting the multiple spatial spectral peaks into a cost function for calculation to obtain the i-th cost function value; continuing to use the (i + 1)-th steering vector and the preprocessed echo signal to determine the (i + 1)-th cost function value, where the (i + 1)-th steering vector is calculated using the (i + 1)-th group of multi-dimensional error particles, until the I-th cost function value is obtained; selecting the maximum value among the I cost function values, and using the spatial spectral peak corresponding to the maximum value as the multiple target angle values.

[0008] In some embodiments, the i-th group of multi-dimensional error particles includes: array position error particles, array velocity error particles, and course velocity error particles.

[0009] In some embodiments, preprocessing a received set of echo signals with array manifold distortion to obtain preprocessed echo signals includes: performing pulse compression processing on the echo signals to obtain pulse compression signals; performing envelope alignment processing on the pulse compression signals to obtain the preprocessed echo signals.

[0010] In some embodiments, performing feature extraction on the preprocessed echo signal to determine multiple strong scattering points includes: performing autofocus processing on the preprocessed echo signal to obtain an autofocus processed signal; performing azimuth Fourier transform on the autofocus processed signal to obtain an ISAR two-dimensional image; performing feature extraction on the ISAR two-dimensional image to obtain the multiple strong scattering points.

[0011] In some embodiments, performing feature extraction on the ISAR two-dimensional image to obtain the multiple strong scattering points includes: obtaining a preset strong scattering point extraction value; performing feature extraction on the ISAR two-dimensional image to obtain multiple feature points; using the preset strong scattering point extraction value to extract the feature points greater than or equal to the preset strong scattering point extraction value among the multiple feature points to obtain the multiple strong scattering points.

[0012] In some embodiments, calculating an array ISAR three-dimensional image using the multiple strong scattering points and the multiple target angle values includes: calculating corresponding three-dimensional coordinates of multiple scattering points using the multiple strong scattering points and the multiple target angle values; obtaining the array ISAR three-dimensional image based on the three-dimensional coordinates of the multiple scattering points.

[0013] In some embodiments, the expression of the ISAR two-dimensional image is:

[0014]

[0015] where G is the ISAR two-dimensional image, σ is the backscattering coefficient, B is the signal bandwidth, t r is the fast time in the range direction, c is the propagation speed of electromagnetic waves, t m is the slow time in the azimuth direction, Δf a = 1 / T a , T a is the preset coherent processing interval, (x, y) is the target position coordinates at the initial moment, ω e is the effective rotation speed, and λ is the wavelength.

[0016] In some embodiments, the expression of the i-th steering vector is:

[0017]

[0018] where is the i-th steering vector, is the array direction error, is the course error, u = (u1, u2,..., u N ) T , is the array direction position error particle in the i-th group of multi-dimensional error particles, v = (v1, v2,..., v N ) T , is the array direction velocity error particle in the i-th group of multi-dimensional error particles, w = (w1, w2,..., w N ) T , is the course velocity error particle in the i-th group of multi-dimensional error particles, θ k is the incoming wave direction angle, and j is the imaginary part.

[0019] In some embodiments, the expression of the i-th cost function value satisfies:

[0020]

[0021] where A s i is the sum of the maximum peaks in the spatial spectrum within the i-th consecutive pulse, Δθ i is the change amount of the maximum peak position within the i-th consecutive pulse, PCR i is the ratio of the maximum value of the i-th spurious peak to the maximum peak of the scattering point, σ i is the standard deviation of the i-th scattering point position. h1 and h2 are A s i and Δθ iThe corresponding weights, where h3 and h4 are the PCRs i and σ i The corresponding weights.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0023] The proposed array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform effectively solves the problem of degraded imaging quality caused by array manifold distortion in complex dynamic scenarios through multi-step collaborative innovation, significantly improving the three-dimensional reconstruction accuracy and system robustness. The technical solution first preprocesses the echo signal through pulse compression and envelope alignment to suppress the phase shift caused by motion; then obtains the ISAR two-dimensional image using autofocus and azimuth Fourier transform, extracts the strong scattering points in the ISAR two-dimensional image, combines the multi-dimensional error particles generated by the improved PSO algorithm to dynamically simulate the actual motion distortion, and realizes high-resolution DOA estimation based on the SBL algorithm, adaptively correcting the steering vector to compensate for the array manifold deviation; finally, fuses the range-Doppler information of the strong scattering points and the multi-view angle estimation, and generates a three-dimensional image through spatial geometric calculation. The first solution proposed based on the present invention can significantly improve the imaging accuracy through dynamic error modeling, the second is to suppress multipath coherent signals through SBL to improve the strong anti-interference ability; the third is to shorten the single-frame processing time through the improved PSO algorithm combined with GPU acceleration to meet the real-time imaging requirements, with less hardware resource occupancy and can be deployed on a moving platform. This solution combines high precision, strong environmental adaptability and real-time performance, provides a reliable three-dimensional perception means for complex scenarios, and can provide technical migration value for fields such as communication and medical imaging. Brief Description of the Drawings

[0024] Figure 1 is a schematic diagram of the antenna array model applied to the array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform provided by an embodiment of the present invention;

[0025] Figure 2 is a schematic flowchart of the array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform provided by an embodiment of the present invention;

[0026] Figure 3 is a schematic diagram of the simulation model provided by an embodiment of the present invention;

[0027] Figure 4 is a schematic diagram of the simulation ArISAR system provided by an embodiment of the present invention;

[0028] Figure 5 is a simulation diagram of the ArISAR three-dimensional imaging result when only array-direction errors exist provided by an embodiment of the present invention;

[0029] Figure 6It is a simulation diagram of the ArISAR three-dimensional imaging result when only the heading error exists provided by the embodiment of the present invention;

[0030] Figure 7 It is a simulation diagram of the ArISAR three-dimensional imaging result when there are array direction errors and heading errors provided by the embodiment of the present invention. Specific embodiments

[0031] The following further describes the present invention in detail with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

[0032] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present invention, "a plurality" means two or more unless otherwise specifically defined.

[0033] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine the different embodiments or examples described in this specification.

[0034] Now, with reference to the accompanying drawings, the array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform provided by the embodiment of the present invention will be introduced in detail.

[0035] In a possible application scenario, the array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform provided by the present invention is applied to Figure 1 the antenna array model in the non-ideal trajectory scenario shown.

[0036] Figure 2 It is a schematic flow diagram of the array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform provided by the embodiment of the present invention. As Figure 2 shown, the method includes:

[0037] Step 110: Obtain a set of echo signals with array manifold distortion.

[0038] Here, in a non-ideal motion scenario, the ArISAR is used to obtain the echo signal. Due to environmental factors (temperature, humidity, and multipath effects), hardware problems (such as channel inconsistencies between different array elements, including differences in amplifier gain, filter characteristics, or sampling biases of the ADC, which can all cause the actual received signal to deviate from the theoretical model; also, changes in the operating frequency of the radar system, frequency offsets, etc.), there are array manifold distortions in the received echo signal. Array manifold distortions will cause the original amplitude and phase offsets in the signal, resulting in the actual array response not matching the ideal model, thereby affecting the performance of the radar system. Here, the motion trajectory under non-ideal motion refers to the trajectory of an object during actual motion, which deviates from the ideal model (such as the theoretical trajectory under assumptions of no friction, no air resistance, and infinite control precision) due to mechanical errors or external disturbances during the actual motion of the object.

[0039] In a possible implementation, taking the projection of the radar on the horizontal plane as the coordinate origin, with the motion direction of the target (moving platform) as the y-axis and the direction perpendicular to the horizontal plane as the z-axis, a three-dimensional coordinate system is established. Along the x-axis direction, N radar array elements are evenly distributed with a spacing of d, and each array element sails at a constant speed v along the y-axis direction. u The center coordinates of the ship target in the O-XYZ coordinate system at the starting time are (x0, y0, z0), and the target moves at a constant speed v along the y-axis direction. After selecting the time period, ignoring the swaying motion of the target, the target motion is approximated as translational motion. The working mode of the array set in this paper is the one-transmit-multiple-receive mode. R0 is the distance from the transmitting array element to the target, and the echo signals at each array element are as follows: s Among them, t is the fast time in the range direction, t r is the slow time, σ is the echo amplitude, p(t m ) is the echo envelope, λ is the radar wavelength, R r represents the distance from the target to the transmitting array element at the time point of slow time t t , and R m represents the distance from the target to the receiving array element at the time point of slow time t r . m

[0040] Among them, taking the transmitting array element E1 as a reference, the position of this array element at time t m can be expressed as: The position of the ship at time t m can be expressed as:

[0041]

[0042] Furthermore, R t can be expressed as:

[0043]

[0044] Step 120: Preprocess the echo signal to obtain the preprocessed echo signal.

[0045] Here, step 120 specifically includes: performing pulse compression processing on the echo signal to obtain a pulse-compressed signal; performing envelope alignment processing on the pulse-compressed signal to obtain the preprocessed echo signal.

[0046] Through pulse compression processing, the range resolution can be effectively improved without reducing the detection range, and envelope alignment processing can correct the echo envelope offset caused by target movement or platform vibration, ensuring the phase consistency of multi-pulse signals, thereby achieving high-resolution imaging.

[0047] Step 130: Extract features from the preprocessed echo signal to determine multiple strong scattering points.

[0048] Specifically, step 130 includes: performing autofocus processing on the preprocessed echo signal to obtain the signal after autofocus processing; performing azimuth Fourier transform on the signal after autofocus processing to obtain an ISAR two-dimensional image; performing feature extraction on the ISAR two-dimensional image to obtain multiple strong scattering points.

[0049] Here, using the multi-prominent point method, the initial phase of the data is corrected, that is, autofocus is performed, to obtain the echo signal after translational compensation. After azimuth Fourier transform, an ISAR two-dimensional image in range time domain - azimuth frequency domain is obtained. The expression of this ISAR two-dimensional image is:

[0050]

[0051] where G is the ISAR two-dimensional image, σ is the backscattering coefficient, B is the signal bandwidth, t r is the fast time in the range direction, c is the electromagnetic wave propagation speed, t m is the slow time in the azimuth direction, Δf a = 1 / T a , T a is the preset coherent processing interval, (x, y) is the target position coordinate at the initial moment, ω e is the effective rotation speed, and λ is the wavelength.

[0052] Here, performing feature extraction on the ISAR two-dimensional image to obtain multiple strong scattering points includes: obtaining a preset strong scattering point extraction value; performing feature extraction on the ISAR two-dimensional image to obtain multiple feature points; using the preset strong scattering point extraction value to extract the feature points greater than or equal to the preset strong scattering point extraction value among the multiple feature points to obtain multiple strong scattering points.

[0053] Here, the amplitude threshold of the extracted strong scattering points is set as A min, the strong scattering points extracted from the ISAR two-dimensional image G are expressed as: where is the scattering point coordinate, is the range cell number, is the Doppler frequency number.

[0054] Step 140: Combine multiple multi-dimensional error particles to perform DOA estimation on the preprocessed echo signal to determine multiple target angle values. Among them, the multiple multi-dimensional error particles are obtained by using an improved PSO algorithm and are used to simulate the array manifold distortion in the echo signal.

[0055] Here, the improved PSO algorithm is first introduced. The PSO algorithm, whose full English name is Particle Swarm Optimization and Chinese name is Particle Swarm Algorithm, is an optimization algorithm based on swarm intelligence. Traditional PSO algorithms may have some problems, such as being prone to falling into local optima, especially when dealing with high-dimensional, multi-modal or complex optimization problems. In addition, the selection of parameters has a great impact on the algorithm performance, such as the inertia weight, acceleration constant, etc. In some cases, the convergence speed may not be fast enough, or premature convergence may occur, that is, it prematurely stagnates at a certain local optimal solution and cannot jump out. To overcome the above problems, existing improvement methods include: 1. Dynamically adjust the inertia weight and acceleration constant; 2. Introduce mutation operations to increase diversity; 3. Multiple swarm cooperation and information exchange; 4. Combine local search methods; 5. Adaptive parameter adjustment mechanism; 6. Extend to multi-objective optimization problems.

[0056] In this application, the array azimuth position error particle, the array azimuth velocity error particle and the course velocity error particle are used as update targets. During the update process using the improved PSO algorithm, the position and velocity of each dimension of the particle are restricted within the allowable range. If the current acceleration of the particle causes its velocity V i in a certain dimension to exceed the maximum velocity V dmax of this dimension, to avoid local optima caused by particles piling up near the maximum velocity upper limit, the velocity of this dimension is reset to a random value within the allowable range. Generally speaking, the selection of V dmax should not exceed the particle width range. If V dmax is too large, the particle may fly past the position of the optimal solution; if it is too small, the global search ability of the particle may be reduced.

[0057] In a possible implementation manner, each time the improved PSO algorithm outputs a group of multi-dimensional particles, and each group of multi-dimensional particles includes multiple multi-dimensional error particles. The multiple multi-dimensional error particles include: the array azimuth position error particle, the array azimuth velocity error particle and the course velocity error particle.

[0058] Here, step 140 includes: using the SBL algorithm to process the i-th steering vector and the preprocessed echo signal to obtain the i-th spatial spectrogram, where i is a positive integer and less than or equal to I; among them, the i-th steering vector is calculated using the i-th group of multi-dimensional error particles; traversing the i-th spatial spectrogram to extract multiple spatial spectral peaks; substituting the multiple spatial spectral peaks into the cost function for calculation to obtain the i-th cost function value; continuing to use the (i + 1)-th steering vector and the preprocessed echo signal to determine the (i + 1)-th cost function value, and the (i + 1)-th steering vector is calculated using the (i + 1)-th group of multi-dimensional error particles, until the I-th cost function value is obtained; selecting the maximum value among the I cost function values, and using the spatial spectral peak corresponding to the maximum value as multiple target angle values.

[0059] Here, the SBL algorithm is introduced first. The SBL algorithm, whose full English name is Sparse Bayesian Learning and full Chinese name is Sparse Bayesian Learning, is a sparse signal reconstruction method based on Bayesian statistical inference. It is often used in DOA estimation, has extremely high resolution, does not require the prior knowledge of the number of signal sources, can automatically identify the direction of effective signals, and is robust to coherent signals (such as multipath scattering), and can overcome the limitations of MUSIC / ESPRIT.

[0060] Here, the i-th group of multi-dimensional error particles includes: multiple array position error particles, multiple array velocity error particles, and multiple course velocity error particles. The expression of the i-th steering vector is:

[0061]

[0062] Among them, is the i-th steering vector, is the array direction error, is the course error, u = (u1, u2,..., u N ) T , is the array position error particle in the i-th group of multi-dimensional error particles, v = (v1, v2,..., v N ) T , is the array velocity error particle in the i-th group of multi-dimensional error particles, w = (w1, w2,..., w N ) T , is the course velocity error particle in the i-th group of multi-dimensional error particles, θ k is the incident wave direction angle, and j is the imaginary part.

[0063] And, the expression of the i-th cost function value satisfies: Among them, A s i is the sum of the maximum peaks in the spatial spectrum within the i-th consecutive pulse, Δθi is the change amount of the maximum peak position within the i-th consecutive pulse, PCR i is the ratio of the maximum value of the i-th spurious peak to the maximum peak of the scattering point, σ i is the standard deviation of the position of the i-th scattering point. h1 and h2 are the weights corresponding to A s i and Δθ i respectively, and h3 and h4 are the weights corresponding to PCR i and σ i respectively.

[0064] It should be noted that the calculation expressions of each cost function value are the same; the spurious peaks in the above formula refer to the peaks in the spatial spectrum that are less than the preset value, the scattering point peak can be understood as the peak in the spatial spectrum that is greater than the preset value, and the scattering point position is the angle where the scattering point peak is located.

[0065] By selecting the maximum value among the I cost function values and then using the spatial spectrum peak corresponding to this maximum value as multiple target angle values, accurate DOA estimation of the signal can be performed on the basis of considering the array manifold distortion, realizing the compensation of errors and the angle measurement of the target.

[0066] Step 150: Calculate and obtain the array ISAR three-dimensional image by using multiple strong scattering points and multiple target angle values.

[0067] Here, after obtaining multiple strong scattering points and multiple target angle values, the three-dimensional spatial coordinates of each scattering point are solved by combining the azimuth resolution, the distance from each scattering point to the array element, and the spatial geometric relationship, so as to obtain the array ISAR three-dimensional image.

[0068] Aiming at the problem that existing 3D imaging technologies are not applicable to targets with complex motion trajectories, and the imaging quality is not considered affected by array manifold distortion during imaging, and they do not have universality, the present invention proposes an array ISAR 3D imaging method based on a non-ideal trajectory distributed moving platform. On the non-ideal moving platform, this method uses an array radar to collect echo signals with array manifold distortion. Subsequently, pulse compression is used to improve the range resolution, and envelope alignment is used to correct the phase shift caused by platform motion to ensure signal consistency, completing the preprocessing of the echo signals. Then, autofocus and azimuth FFT transformation are respectively performed on the preprocessed signals to obtain an ISAR 2D image. Feature extraction of strong scatterers, DOA estimation and error compensation are performed on the ISAR 2D image. Finally, the 3D coordinates are generated using the positions of the extracted strong scatterers in the 2D image and the DOA estimation values to construct an ISAR 3D image. Based on the method provided by the present invention, it is possible to combine the ISAR 2D image and array multi-view observation, break through the single-view resolution limit, achieve true 3D imaging, effectively improve the imaging quality, and be applicable to various dynamic scenarios. Moreover, through dynamic error modeling and multi-source information fusion, the key problems of array ISAR imaging under non-ideal motion are solved, and it has high precision, strong robustness and engineering practicability.

[0069] To verify the above technical effects, a simulation software was used for verification. First, a Figure 3 simulation model as shown was established. This simulation model contains 53 scatterers, and the appearance size is 40m×100m×40m. It is assumed that the scattering coefficient of each scatterer in the model is 1. The main simulation parameters are shown in Table 1. The simulation experiment uses the Figure 4 ArISAR system as shown. To verify the effectiveness and robustness of the proposed algorithm, three antenna error configuration scenarios were set respectively. Scenario 1: Only the array direction error scenario; Scenario 2: Only the course error scenario; Scenario 3: Both the array direction error and the course error exist. The obtained ArISAR 3D imaging results are as shown in the appendix Figures 5 - 7 as shown. To quantitatively describe the performance of the algorithm, the root mean squared error (RMSE) between the reconstructed scatterer coordinates and the ideal 3D geometric model was used as the quantization index in the experiment, and its expression is: where, is the estimated 3D coordinates of the scatterer, (x p , y p , z p ) are the actual 3D coordinates of the scatterer, and P is the total number of scatterers.

[0070] At the same time, the relative error RE between the estimated error result and the actual error is used as a quantization index, and its expression is: RE = (xrec - xref) / xref; where xrec is the reconstructed coordinate and xref is the actual coordinate.

[0071] Table 1

[0072] parameter value parameter value element spacing 1m number of elements eight carrier frequency 10 GHz bandwidth 300 MHz PRF 300 Hz target distance 4 km pulse width 1 μs sampling frequency 400 MHz number of range cells 1300 number of azimuth cells 300

[0073] Combined with Figures 5 - 7 The RMSE between the imaging result and the model, the error parameter estimation result, and the relative error between the error estimation result and the actual error are given. Combining the experimental results shown Figures 5 - 7 and the data indexes therein, it can be concluded that the present invention shows excellent performance in different non-ideal trajectory scenarios, proving the effectiveness and strong robustness of the proposed method.

[0074] The above content is a further detailed description of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is only limited to these descriptions. For those of ordinary skill in the technical field to which the present invention belongs, without departing from the concept of the present invention, several simple deductions or substitutions can be made, and all should be regarded as belonging to the protection scope of the present invention.

Claims

1. An array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform, characterized in that Including: Obtaining a set of echo signals with array manifold distortion; Preprocessing the echo signals to obtain preprocessed echo signals; Performing feature extraction on the preprocessed echo signals to determine multiple strong scattering points; Combining multiple multi-dimensional error particles to perform DOA estimation on the preprocessed echo signals to determine multiple target angle values, where the multiple multi-dimensional error particles are obtained by using an improved PSO algorithm and are used to simulate the array manifold distortion in the echo signals; Calculating and obtaining an array ISAR three-dimensional image by using the multiple strong scattering points and the multiple target angle values.

2. The array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform according to claim 1, wherein, The combining multiple multi-dimensional error particles to perform DOA estimation on the preprocessed echo signals to determine multiple target angle values includes: Using the SBL algorithm to process by combining the i-th steering vector and the preprocessed echo signals to obtain the i-th spatial spectrogram, where i is a positive integer and less than or equal to I; the i-th steering vector is obtained by using the i-th group of multi-dimensional error particles; Traversing the i-th spatial spectrogram to extract multiple spatial spectrum peaks; Substituting the multiple spatial spectrum peaks into a cost function for calculation to obtain the i-th cost function value; continuing to use the (i + 1)-th steering vector and the preprocessed echo signals to determine the (i + 1)-th cost function value, where the (i + 1)-th steering vector is obtained by using the (i + 1)-th group of multi-dimensional error particles, until the I-th cost function value is obtained; Selecting the maximum value among the I cost function values and using the spatial spectrum peak corresponding to the maximum value as the multiple target angle values.

3. The array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform according to claim 2, wherein, The i-th group of multi-dimensional error particles includes: array position error particles, array velocity error particles, and course velocity error particles.

4. The array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform according to claim 1, wherein The preprocessing the received set of echo signals with array manifold distortion to obtain preprocessed echo signals includes: Performing pulse compression processing on the echo signals to obtain pulse compression signals; Performing envelope alignment processing on the pulse compression signals to obtain the preprocessed echo signals.

5. The array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform according to claim 1, wherein The performing feature extraction on the preprocessed echo signals to determine multiple strong scattering points includes: Performing autofocus processing on the preprocessed echo signals to obtain autofocus processed signals; Performing azimuth Fourier transform on the autofocus processed signals to obtain an ISAR two-dimensional image; Performing feature extraction on the ISAR two-dimensional image to obtain the multiple strong scattering points.

6. The array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform according to claim 5, wherein The performing feature extraction on the ISAR two-dimensional image to obtain the multiple strong scattering points includes: Obtaining a preset strong scattering point extraction value; Performing feature extraction on the ISAR two-dimensional image to obtain multiple feature points; Using the preset strong scattering point extraction value to extract the feature points greater than or equal to the preset strong scattering point extraction value among the multiple feature points to obtain the multiple strong scattering points.

7. The array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform according to claim 1, characterized in that The calculating and obtaining an array ISAR three-dimensional image by using the multiple strong scattering points and the multiple target angle values includes: Calculating the three-dimensional coordinates of the corresponding multiple scattering points by using the multiple strong scattering points and the multiple target angle values; Based on the three-dimensional coordinates of the multiple scattering points, the three-dimensional image of the array ISAR is obtained.

8. The array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform according to claim 5, characterized in that, The B expression of the two-dimensional ISAR image is as follows: Where, G is the ISAR two-dimensional image, σ is the backscattering coefficient, is the signal bandwidth, t r is the fast time in the range direction, c is the propagation speed of electromagnetic waves, t m is the slow time in the azimuth direction, Δf a = 1 / T a , is the preset coherent processing interval, (x, y) is the target position coordinate at the initial moment, ω e is the effective rotation speed, and λ is the wavelength.

9. The array ISAR three-dimensional imaging method based on a non-ideal trajectory distributed moving platform according to claim 5, characterized in that The expression of the i-th steering vector is: Among them, is the i-th steering vector, is the array direction error, is the course error, u = (u1, u2, …, u N ) T , is the array direction position error particle in the i-th group of multi-dimensional error particles, v = (v1, v2, …, v N ) T , is the array direction velocity error particle in the i-th group of multi-dimensional error particles, w = (w1, w2, …, w N ) T , is the course velocity error particle in the i-th group of multi-dimensional error particles, θ k is the incident wave direction angle, and j is the imaginary part.

10. The method for three-dimensional imaging of an array ISAR based on a non-ideal trajectory distributed moving platform according to claim 2, wherein The expression of the i-th cost function value satisfies: Among them, A s i is the sum of the maximum peaks in the spatial spectrum within the i-th consecutive pulse, and Δθ i is the change amount of the maximum peak position within the i-th consecutive pulse, PCR i is the ratio of the maximum value of the i-th spurious peak to the maximum peak of the scattering point, σ i is the standard deviation of the position of the i-th scattering point, and h1 and h2 are the weights corresponding to A s i and Δθ i respectively, and h3 and h4 are the weights corresponding to PCR i and σ i respectively.

Images