Method and device for reconstructing moving target trajectory of wide-angle staring SAR based on Gauss-Newton method

Abstract

A method and device for reconstructing moving target trajectories for a wide-angle staring SAR based on the Gauss-Newton method include performing sub-aperture division on the original radar echo signal; performing two-dimensional compression on the original echo signal of each channel in the range and azimuth dimensions to obtain the range-Doppler domain echo of each channel; and preprocessing the range-Doppler domain echo in the range-Doppler domain; extracting the phase history and envelope of the moving target based on the preprocessed range-Doppler domain echo to estimate the radial velocity of the moving target; using a maximum likelihood algorithm to estimate each order phase parameter in the extracted phase history; obtaining the motion parameters of the moving target based on the relationship between each order phase parameter and the motion parameters; obtaining a rough motion trajectory reconstruction result based on the motion parameters of the moving target; segmenting the rough motion trajectory of the moving target to obtain the precise reconstructed position of the moving target in each segment; and splicing the reconstructed positions to obtain a complete motion trajectory reconstruction result.

Description

Technical Field

[0001] The present invention belongs to the field of radar signal processing, and in particular relates to a method and device for reconstructing moving target trajectories of wide-angle staring SAR based on the Gauss-Newton method. Background Art

[0002] Wide-angle Staring Synthetic Aperture Radar (WasSAR) is a special mode of Synthetic Aperture Radar (SAR). In this mode, the radar illuminates the region of interest (ROI) for a long time, accumulating a large coherent accumulation angle. The main implementation of wide accumulation angles is spotlight SAR, such as circular SAR (CSAR) and strip spotlight SAR. Combining WasSAR with Ground Moving Target Indication (GMTI) technology can achieve long-term continuous monitoring of the ROI. It is currently widely used in military and civilian fields, such as battlefield surveillance and traffic monitoring. Geolocation of moving targets is an important practical application of the WasSAR-GMTI system and can generally be divided into two categories: shadow-based geolocation and Doppler energy-based geolocation.

[0003] The shadow of a moving target will be projected at the actual location without any offset, which is effective for the geolocation of moving targets. The most common approach is to detect the shadow of a moving target through morphological operations. The establishment of a background model is the core step of shadow detection. However, the geolocation method based on shadows is limited by the quality of shadows. Due to the low signal-to-noise ratio of SAR images, the fast speed of target movement, and obstruction by objects (such as buildings and trees), the shadow of a moving target is prone to blurring and smearing. The Doppler energy detection method based on the range-doppler (RD) domain has better robustness and stability in the geolocation of moving targets as the information dimension increases.

[0004] The primary task of geolocation is to resolve the Doppler shift, determined by both the azimuth position and radial velocity of a moving target—that is, the positional uncertainty. Currently, a common approach involves using multiple perspectives to establish the equation of motion and solve for the motion parameters. However, due to the large condition number, the parameter estimation equations are ill-conditioned. To solve the ill-conditioned equations, the moving target is assumed to have a constant velocity over a long observation period. However, over extended observation periods, the target may undergo nonlinear motion and rotation. Therefore, this approach is not applicable to field data processing. Alternatively, using an on-track multi-channel system, with the increased spatial degrees of freedom, radial velocity can be estimated through interferometric phase estimation. The additional Doppler shift associated with radial velocity can be directly determined. However, this method is sensitive to velocity estimation and is easily affected by clutter and noise in the actual data. To eliminate geolocation errors, traditional methods search for optimal results by multidimensional parameter maximization. However, this multidimensional search requires a high computational load, and the ill-conditioned equations can easily lead the results to local minima. Other methods utilize a priori road information, but this requires prior scene information, limiting their practical application. Summary of the Invention

[0005] In view of the shortcomings of the prior art, the present invention provides a method and device for reconstructing moving target trajectories for wide-angle staring SAR based on the Gauss-Newton method.

[0006] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] In one aspect, the present invention provides a method for reconstructing moving target trajectories of wide-angle staring SAR based on the Gauss-Newton method, comprising:

[0008] (1) Perform sub-aperture division on the original radar echo signal to obtain multi-channel original echo signals;

[0009] (2) Perform two-dimensional compression on the original echo signals of each channel in the range and azimuth dimensions to obtain the range-Doppler domain echo of each channel;

[0010] (3) performing preprocessing in the range-Doppler domain based on the range-Doppler domain echo of each channel, wherein the preprocessing includes multi-channel registration and clutter suppression, to obtain the range-Doppler domain echo after preprocessing;

[0011] (4) Based on the pre-processed range-Doppler domain echo, the phase history and envelope of the moving target are extracted to estimate the radial velocity of the moving target;

[0012] (5) The phase parameters of the moving target phase history are coupled with the carrier position and the motion information of the moving target. The maximum likelihood algorithm is used to estimate the phase parameters of each order in the extracted phase history.

[0013] (6) Based on the relationship between each order phase parameter and motion parameter, the Gauss-Newton method is used to decouple the motion parameters and obtain the motion parameters of the moving target;

[0014] (7) Obtaining a rough motion trajectory reconstruction result based on the motion parameters of the moving target;

[0015] (8) Segment the rough motion trajectory of the moving target, and use the Gauss-Newton method to decouple the extracted envelope from the rough motion trajectory of each segment to obtain the precise reconstructed position of the moving target in each segment;

[0016] (9) The reconstructed positions of the moving targets in each segment of the motion trajectory are spliced ​​together to obtain the complete motion trajectory reconstruction result.

[0017] On the other hand, the present invention provides a wide-angle staring SAR moving target trajectory reconstruction device based on the Gauss-Newton method, comprising:

[0018] The original echo processing module is used to perform sub-aperture division on the original radar echo signal to obtain multi-channel original echo signals; perform two-dimensional compression on the original echo signal of each channel in the range and azimuth dimensions to obtain the range-Doppler domain echo of each channel; perform preprocessing in the range-Doppler domain based on the range-Doppler domain echo of each channel, wherein the preprocessing includes multi-channel registration and clutter suppression to obtain the preprocessed range-Doppler domain echo;

[0019] The parameter acquisition module extracts the phase history and envelope of the moving target based on the preprocessed range-Doppler domain echo and estimates the radial velocity of the moving target. The phase parameters of the moving target phase history are coupled with the carrier position and the motion information of the moving target. The maximum likelihood algorithm is used to estimate the phase parameters of each order in the extracted phase history. Based on the relationship between each order phase parameter and the motion parameter, the Gauss-Newton method is used to decouple the motion parameters to obtain the motion parameters of the moving target.

[0020] The position reconstruction module obtains a rough motion trajectory reconstruction result based on the motion parameters of the moving target; the rough motion trajectory of the moving target is segmented, and the extracted envelope is decoupled from the rough motion trajectory of each segment using the Gauss-Newton method to obtain the precise reconstructed position of the moving target in each segment;

[0021] The trajectory reconstruction module is used to splice the reconstructed position of the moving target in each segment of the motion trajectory to obtain a complete motion trajectory reconstruction result.

[0022] The above-mentioned wide-angle staring SAR moving target trajectory reconstruction method and device based on the Gauss-Newton method first divides the original echo signal received by the radar into multiple sub-apertures, and performs two-dimensional compression on the original echo signal of each channel in the range and azimuth dimensions to obtain the range-Doppler domain echo of each channel. After preprocessing the range-Doppler domain echo, the phase history and envelope of the moving target are extracted, the radial velocity of the moving target is estimated, and the polynomial parameters of the phase history are estimated using the maximum likelihood method (Quasi-maximum Likelihood, QML). Secondly, the Gauss-Newton algorithm is used to decouple the motion parameters, and the rough trajectory can be reconstructed using the motion parameters. Subsequently, the motion trajectory of the moving target is segmented modeled using the extracted envelope history of the moving target, and each segment of the motion trajectory is estimated using the corrected envelope history. Finally, after processing each sub-aperture, an accurate and complete trajectory of the moving target is obtained. The present invention can solve the problem that existing methods are difficult to achieve geographical positioning of moving targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the structures shown in these drawings without paying any creative work.

[0024] Figure 1 A schematic flow chart of a method for reconstructing moving target trajectories for wide-angle staring SAR based on the Gauss-Newton method provided in one embodiment;

[0025] Figure 2 A schematic diagram of a geometric model for multi-channel WasSAR data acquisition provided in one embodiment;

[0026] Figure 3 A schematic diagram of segmented modeling of a motion trajectory provided by an embodiment;

[0027] Figure 4 The measured data optical image, imaging results and cooperative vehicle target map provided by an embodiment;

[0028] Figure 5 Position information measured by an airborne GPS provided in an embodiment;

[0029] Figure 6 A comparison chart of the PPS coefficients estimated using the QML algorithm for objective 1 provided in one embodiment and the theoretical values, where (a), (b), and (c) are the first-order coefficients, second-order coefficients, and third-order coefficients, respectively;

[0030] Figure 7Graphs comparing target motion parameters and true values ​​provided in one embodiment, wherein (a), (b), (c), and (d) are the x-direction velocity, y-direction velocity, x-direction acceleration, and y-direction acceleration of target 1, respectively; (e), (f), (g), and (h) are the x-direction velocity, y-direction velocity, x-direction acceleration, and y-direction acceleration of target 2, respectively; and (i), (j), (k), and (l) are the x-direction velocity, y-direction velocity, x-direction acceleration, and y-direction acceleration of target 3, respectively.

[0031] Figure 8 A diagram showing the result of extracting envelope information of target 1 provided in one embodiment;

[0032] Figure 9 Figure 1 is a trajectory reconstruction result diagram provided by an embodiment, wherein Figure (a), Figure (b) and Figure (c) are trajectory reconstruction structure diagrams of target 1, target 2 and target 3 respectively. DETAILED DESCRIPTION

[0033] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0034] In one embodiment, referring to Figure 1 , provides a wide-angle staring SAR moving target trajectory reconstruction method based on the Gauss-Newton method, including:

[0035] (1) Perform sub-aperture division on the original radar echo signal to obtain multi-channel original echo signals;

[0036] (2) Perform two-dimensional compression on the original echo signals of each channel in the range and azimuth dimensions to obtain the range-Doppler domain echo of each channel;

[0037] (3) performing preprocessing in the range-Doppler domain based on the range-Doppler domain echo of each channel, wherein the preprocessing includes multi-channel registration and clutter suppression, to obtain the range-Doppler domain echo after preprocessing;

[0038] (4) Based on the pre-processed range-Doppler domain echo, the phase history and envelope of the moving target are extracted to estimate the radial velocity of the moving target;

[0039] (5) The phase parameters of the moving target phase history are coupled with the carrier position and the motion information of the moving target. The maximum likelihood algorithm is used to estimate the phase parameters of each order in the extracted phase history.

[0040] (6) Based on the relationship between each order phase parameter and motion parameter, the Gauss-Newton method is used to decouple the motion parameters and obtain the motion parameters of the moving target;

[0041] (7) Obtaining a rough motion trajectory reconstruction result based on the motion parameters of the moving target;

[0042] (8) Segment the rough motion trajectory of the moving target, and use the Gauss-Newton method to decouple the extracted envelope from the rough motion trajectory of each segment to obtain the precise reconstructed position of the moving target in each segment;

[0043] (9) The reconstructed positions of the moving targets in each segment of the motion trajectory are spliced ​​together to obtain the complete motion trajectory reconstruction result.

[0044] Reference Figure 2 In one embodiment, in step (1), the original radar echo signal is received by a SAR radar; and step (2) includes the following steps:

[0045] Perform range pulse pressure and range-direction fast Fourier transform on the original echo signals of each channel to obtain the echo fundamental frequency signal of the moving target:

[0046] s e (r,t m )=σg e sinc{2B[rR e (t m ; r p )] / c}·exp[-j4πf c R e (t m ; r p ) / c]

[0047] Among them, σ represents the amplitude of the echo, g e Represents the bidirectional antenna pattern of each channel. B,f c and c represent the bandwidth, carrier frequency and speed of light of the echo respectively. r represents the slant range unit, t m represents slow time; r p Represents the actual position of the moving target; R e (t m ; r p ) represents the slant range from the carrier to the target; j represents the imaginary unit; exp(·) represents the exponential operation with the natural number e as the base;

[0048] The original echo signals of each channel are subjected to fast Fourier transform in the azimuth dimension to achieve azimuth compression and obtain range-Doppler domain echoes.

[0049] Since the antenna spacing causes the time delay difference of each channel observing the same scene, the correlation of the multi-channel received signals is destroyed. Therefore, it is necessary to use the speed of the carrier aircraft and the baseline length between the multiple channels to compensate for the time delay error. Once the compensation is completed, the multi-channel registration is completed. The clutter suppression effect in step (3) is proportional to the spatial degree of freedom. As the spatial degree of freedom increases, the multi-channel clutter suppression effect becomes better and better.

[0050] Traditional multi-channel moving target detection methods include Displaced Phase Center Array (DPCA), Along Track Interferometry (ATI), Clutter Suppression Interference (CSI), and Space-time Adaptive Processing (STAP).

[0051] In step (4), the radial velocity of the moving target can be estimated by methods such as weighted adaptive matched filtering (WAMF) and subspace projection (SP).

[0052] Step (5) comprises the following steps:

[0053] R(t m ; r p ) at the center of the subaperture t m =t k Taylor expansion:

[0054]

[0055] Where R(t m ; r p ) is the slant distance of the moving target from the carrier aircraft, which couples the position of the carrier aircraft and the motion information of the moving target; R′(t k ; r p ), R″(t k ; r p ), R″′(t k ; r p ), R″″(t k ; r p ) and o(t m ; r p ) represent the first-order, second-order, third-order, fourth-order and higher-order Taylor expansion coefficients of the phase history respectively;

[0056] At this time, the echo expression of the moving target can be expressed as

[0057]

[0058] The above equation converts the phase history estimate into a polynomial phase parameter estimate. This polynomial phase parameter can be expanded into a form consisting of the sum of multiple phase parameters. The polynomial phase parameters are estimated using the maximum likelihood algorithm. The QML estimator is a powerful phase parameter estimation technique that can theoretically estimate phase parameters of any order.

[0059] Because the phase parameters of the moving target phase history are coupled with the carrier position and the motion information of the moving target, the Gauss-Newton method is needed to decouple the motion parameters.

[0060] Step (6) comprises the following steps:

[0061] Using the quadratic and cubic Taylor expansion coefficients R″(t k ; r p )、R″′(t k ; r p ), the improved algorithm of Gauss-Newton method is used to decouple the motion parameters. First, the motion parameter estimation model shown below is established

[0062]

[0063] Among them, r a,i , v a,i and θ k,i Represent the flight radius, speed and azimuth of the aircraft in the i-th sub-aperture, which can be obtained through the onboard GPS; x pk,i 、y pk,i Respectively represent the x-coordinate position and y-coordinate position of the moving target; represents the radial velocity of the moving target estimated by the i-th sub-aperture;

[0064] Convert the above motion parameter estimation model into a least squares problem to obtain the motion parameters of the moving target:

[0065]

[0066] where x i Represents the motion parameters of the target within sub-aperture i.

[0067] According to x i The motion parameters of the moving target are obtained as follows:

[0068]

[0069] in, are the x-direction speed and y-direction speed of the moving target respectively; They are the x-direction acceleration and the y-direction acceleration of the moving target respectively.

[0070] Step (7) includes:

[0071] Reconstruct the rough motion trajectory of the moving target based on the estimated motion parameters of the moving target

[0072]

[0073] Among them, x pk,1 、y pk,1 They represent the initial x-coordinate position and initial y-coordinate position of the moving target respectively.

[0074] Since the estimation of motion parameters is prone to fall into local extreme values ​​and there is error accumulation in parameter estimation, the reconstructed rough motion trajectory will deviate from the true value. It is necessary to segment the rough motion trajectory and use the envelope of the moving target to correct the segmented rough motion trajectory to obtain the accurate reconstructed position of the moving target.

[0075] Extracting the envelope of the moving target in step (4) includes the following steps:

[0076] Extract the envelope information of the echo, assuming that the echo is M units to the right along the azimuth

[0077] r e =[r e (1)r e (2)...r e (M)]

[0078] where r e Represents the complete envelope history of the echo;

[0079] By extracting the phase of the peak echo in the range profile, the envelope history can be obtained according to the conversion relationship between phase and distance.

[0080]

[0081] Where Angle[·] represents the phase extraction, λ represents the wavelength, s(r,t m ) represents the echo signal at the peak position in the range profile.

[0082] Step (8) includes:

[0083] The motion state of the moving target is modeled using a segmentation method. crs is the rough trajectory obtained in the previous step. ideal is the actual trajectory, l plane is the platform trajectory; the motion trajectory is evenly divided into K segments according to the time change, each of which contains slow time units Each sub-segment is denoted as l crs,1 ,l crs,2 ,...,l crs,k ,...,l crs,K , the corresponding platform trajectory is recorded as l plane,1 ,l plane,2 ,...,l plane,k ,...,l plane,K . l crs,k and l plane,k It can be expressed as:

[0084]

[0085] In the kth segment, the target moves at a constant speed; define the starting position (x k ,y k ) and speed (v x,k ,v y,k ), then define the target state variable x under uniform linear motion k It is expressed as follows

[0086] x k =[x k ,y k ,v x,k ,v y,k ] T

[0087] Based on the above target state variables, the measurement equation is

[0088] r k =||l plane,k -l k ||2

[0089] where l k Represents the motion state x k The following motion trajectory,

[0090]

[0091] where Δt m Indicates a slow time interval;

[0092] Building the first expression

[0093] g1(x k )=r k -r e,k

[0094] where r e,k =r e | m∈[1+(k-1)Q,kQ] ;

[0095] In the connection of each sub-segment, a prediction mechanism is introduced to calculate the time unit n after the end of sub-segment k-1. p The target movement trend within the subsegment k is the first n p A known trajectory within a time unit,

[0096]

[0097] Among them, l pre,k is the time unit n after the end of subsegment k-1 p The target movement trend within the target; and establish the second relationship

[0098]

[0099] The target position information and velocity information provided by the rough trajectory are introduced into the measurement equation. The target position information is

[0100] g3(x k )=||l crs,k -l k ||2

[0101] According to the target speed information, the equation is established as

[0102]

[0103] Using the equations established above, the problem of solving the state parameters of the moving target is transformed into a least squares problem

[0104]

[0105] The Gauss-Newton method is used to solve the above least squares problem and obtain the state of the target in each trajectory segment.

[0106] The reconstructed positions of the moving targets in each segment of the motion trajectory are spliced ​​together to obtain the complete motion trajectory reconstruction result.

[0107] At this point, the wide-angle staring SAR moving target trajectory reconstruction method based on the Gauss-Newton method is completed.

[0108] In one embodiment, measured WasSAR data is used, and comparative experiments are designed to illustrate the effectiveness and superiority of the algorithm.

[0109] In one embodiment, the measured WasSAR raw data is recorded by a Ku-band airborne radar, with an aircraft speed of approximately 50 m / s and a flight altitude of approximately 2000 m. Figure 4 、 Figure 5 , Figure 4 The measured data optical image, imaging results and cooperative vehicle target map provided by an embodiment; Figure 5This is location information measured by an airborne GPS provided by an embodiment.

[0110] After performing echo range and azimuth compression, multi-channel registration, and clutter suppression, the QML algorithm is used to estimate the target's phase coefficient. Figure 6 , use QML algorithm to estimate the third-order phase coefficient of target 1. Then use LM algorithm to decouple the motion parameters, refer to Figure 7 , Figure 7 A comparison diagram of target motion parameters and true values ​​provided in an embodiment, wherein Figures (a), (b), (c) and (d) are respectively the x-direction velocity, y-direction velocity, x-direction acceleration and y-direction acceleration of target 1; Figures (e), (f), (g) and (h) are respectively the x-direction velocity, y-direction velocity, x-direction acceleration and y-direction acceleration of target 2; Figures (i), (j), (k) and (l) are respectively the x-direction velocity, y-direction velocity, x-direction acceleration and y-direction acceleration of target 3. Figure 8 , Figure 8 This is a diagram showing the result of extracting the envelope information of target 1 provided by an embodiment. Figure 9 , Figure 9 The trajectory reconstruction result diagram provided by an embodiment, wherein Figure (a), Figure (b) and Figure (c) are the trajectory reconstruction structure diagrams of target 1, target 2 and target 3 respectively. Figure 9 It can be seen that the trajectory reconstruction results of the method proposed in this paper are closer to the true value, which illustrates the effectiveness of the method proposed in this paper.

[0111] The comparison of the positioning accuracy of moving targets between the traditional positioning method and the algorithm proposed in this invention is shown in Table 1.

[0112] Table 1 Positioning error of each algorithm

[0113]

[0114]

[0115] From Table 1, it can be seen that compared with the traditional method, the trajectory reconstruction accuracy of the method of the present invention is higher. Compared with the traditional method, the method provided by the present invention has superiority.

[0116] In one embodiment, a device for reconstructing moving target trajectories for wide-angle staring SAR based on the Gauss-Newton method is provided, comprising:

[0117] The original echo processing module is used to perform sub-aperture division on the original radar echo signal to obtain multi-channel original echo signals; perform two-dimensional compression on the original echo signal of each channel in the range and azimuth dimensions to obtain the range-Doppler domain echo of each channel; perform preprocessing in the range-Doppler domain based on the range-Doppler domain echo of each channel, wherein the preprocessing includes multi-channel registration and clutter suppression to obtain the preprocessed range-Doppler domain echo;

[0118] The parameter acquisition module extracts the phase history and envelope of the moving target based on the preprocessed range-Doppler domain echo and estimates the radial velocity of the moving target. The phase parameters of the moving target phase history are coupled with the carrier position and the motion information of the moving target. The maximum likelihood algorithm is used to estimate the phase parameters of each order in the extracted phase history. Based on the relationship between each order phase parameter and the motion parameter, the Gauss-Newton method is used to decouple the motion parameters to obtain the motion parameters of the moving target.

[0119] The position reconstruction module obtains a rough motion trajectory reconstruction result based on the motion parameters of the moving target; the rough motion trajectory of the moving target is segmented, and the extracted envelope is decoupled from the rough motion trajectory of each segment using the Gauss-Newton method to obtain the precise reconstructed position of the moving target in each segment;

[0120] The trajectory reconstruction module is used to splice the reconstructed position of the moving target in each segment of the motion trajectory to obtain a complete motion trajectory reconstruction result.

[0121] Matters not covered by the present invention are known technologies.

[0122] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0123] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and such modifications and improvements are intended to fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

[0124] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that the present invention is susceptible to various modifications and variations. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the present invention.

Claims

1. A method for reconstructing moving target trajectories for wide-angle staring SAR based on the Gauss-Newton method, characterized in that: include: (1) Perform sub-aperture division on the original radar echo signal to obtain a multi-channel original echo signal; (2) Perform two-dimensional compression on the original echo signals of each channel in the range and azimuth dimensions to obtain the range-Doppler domain echo of each channel; (3) performing preprocessing in the range-Doppler domain based on the range-Doppler domain echo of each channel, wherein the preprocessing includes multi-channel registration and clutter suppression, to obtain the range-Doppler domain echo after preprocessing; (4) Based on the pre-processed range-Doppler domain echo, the phase history and envelope of the moving target are extracted to estimate the radial velocity of the moving target; (5) The phase parameters of the moving target phase history are coupled with the carrier position and the motion information of the moving target. The maximum likelihood algorithm is used to estimate the phase parameters of each order in the extracted phase history. (6) Based on the relationship between each order phase parameter and motion parameter, the Gauss-Newton method is used to decouple the motion parameters and obtain the motion parameters of the moving target; (7) Obtaining a rough motion trajectory reconstruction result based on the motion parameters of the moving target; (8) Segment the rough motion trajectory of the moving target, and use the Gauss-Newton method to decouple the extracted envelope from the rough motion trajectory of each segment to obtain the precise reconstructed position of the moving target in each segment; (9) The reconstructed positions of the moving targets in each segment of the motion trajectory are spliced ​​together to obtain the complete motion trajectory reconstruction result.

2. The method for reconstructing moving target trajectories of wide-angle staring SAR based on the Gauss-Newton method according to claim 1, wherein: The step (2) of echoing includes the following steps: Perform range pulse pressure and range-direction fast Fourier transform on the original echo signals of each channel to obtain the echo fundamental frequency signal of the moving target: s e (r,t m )=σg e sinc{2B[r-R e (t m ;r p )] / c}·exp[-j4πf c R e (t m ;r p ) / c] Among them, σ represents the amplitude of the echo, g e Represents the bidirectional antenna pattern of each channel; B, f c and c represent the bandwidth, carrier frequency and speed of light of the echo respectively; r represents the slant range unit, t m represents slow time; r p Represents the actual position of the moving target; R e (t m ; r p ) represents the slant range from the carrier to the target; j represents the imaginary unit; exp(·) represents the exponential operation with the natural number e as the base; The original echo signals of each channel are subjected to fast Fourier transform in the azimuth dimension to achieve azimuth compression and obtain range-Doppler domain echoes.

3. The method for reconstructing moving target trajectories of wide-angle staring SAR based on Gauss-Newton method according to claim 1, wherein: In the step (3), the clutter suppression effect is proportional to the spatial degree of freedom.

4. The method for reconstructing moving target trajectories of wide-angle staring SAR based on the Gauss-Newton method according to claim 1, wherein: The step (5) comprises the following steps: R(t m ; r p ) at the subaperture center t m =t k Taylor expansion: Where R(t m ; r p ) is the slant distance of the moving target from the carrier aircraft, which couples the position of the carrier aircraft and the motion information of the moving target; R′(t k ; r p ), R″(t k ; r p ), R″′(t k ; r p ), R″″(t k ; r p ) and o(t m ; r p ) represent the first-order, second-order, third-order, fourth-order and higher-order Taylor expansion coefficients of the phase history respectively; At this time, the echo expression of the moving target can be expressed as The above formula converts the phase history estimate into a polynomial phase parameter estimate. The polynomial phase parameter can be expanded into a form of adding multiple phase parameters, and the polynomial phase parameter is estimated using the maximum likelihood algorithm.

5. The method for reconstructing moving target trajectories of wide-angle staring SAR based on Gauss-Newton method according to claim 4, wherein: The step (6) comprises the following steps: Using the quadratic and cubic Taylor expansion coefficients R″(t k ; r p )、R″′(t k ; r p ), the improved algorithm of Gauss-Newton method is used to decouple the motion parameters. First, the motion parameter estimation model shown below is established Among them, r a,i , v a,i and θ k,i Respectively represent the flight radius, speed and azimuth of the aircraft in the i-th sub-aperture, obtained through the onboard GPS; x pk,i 、y pk,i Respectively represent the x-coordinate position and y-coordinate position of the moving target; represents the radial velocity of the moving target estimated by the i-th sub-aperture; Convert the above motion parameter estimation model into a least squares problem to obtain the motion parameters of the moving target: where x i Represents the motion parameters of the target within sub-aperture i.

6. The method for reconstructing moving target trajectories of wide-angle staring SAR based on the Gauss-Newton method according to claim 5, wherein: According to x i The motion parameters of the moving target are obtained as follows: in, are the x-direction speed and y-direction speed of the moving target respectively; They are the x-direction acceleration and the y-direction acceleration of the moving target respectively.

7. The method for reconstructing moving target trajectories of wide-angle staring SAR based on Gauss-Newton method according to claim 1, wherein: The step (7) comprises: Reconstruct the rough motion trajectory of the moving target based on the estimated motion parameters of the moving target Among them, x pk,1 、y pk,1 They represent the initial x-coordinate position and initial y-coordinate position of the moving target respectively.

8. The method for reconstructing moving target trajectories of wide-angle staring SAR based on Gauss-Newton method according to claim 1, wherein: Extracting the envelope of the moving target in step (4) includes the following steps: Extract the envelope information of the echo, assuming that the echo is M units to the right along the azimuth r e ＝[r e (1)r e (2)...r e (M)] where r e Represents the complete envelope history of the echo; By extracting the phase of the peak echo in the range profile, the envelope history can be obtained according to the conversion relationship between phase and distance. Where Angle[·] represents the phase extraction, λ represents the wavelength, s(r,t m ) represents the echo signal at the peak position in the range profile.

9. The method for reconstructing moving target trajectories of wide-angle staring SAR based on Gauss-Newton method according to claim 1, wherein: The step (8) comprises: The motion state of the moving target is modeled using a segmentation method. crs is the rough trajectory obtained in the previous step; l ideal is the actual trajectory, l plane is the platform trajectory; the motion trajectory is evenly divided into K segments according to the time change, where each segment contains a slow time unit Q = M / K; each sub-segment is recorded as l crs,1 ,l crs,2 ,...,l crs,k ,...,l crs,K , the corresponding platform trajectory is recorded as l plane,1 ,l plane,2 ,...,l plane,k ,...,l plane,K ; l crs,k and l plane,k It can be expressed as: In the kth segment, the target moves at a constant speed; define the starting position (x k ,y k ) and speed (v x,k ,v y,k ), then define the target state variable x under uniform linear motion k It is expressed as follows x k =[x k ,y k ,v x,k ,v y,k ] T Based on the above target state variables, the measurement equation is r k =||l plane,k -l k ||2 where l k Represents the motion state x k The following motion trajectory, where Δt m Indicates a slow time interval; Building the first expression g1(x k )=r k -r e,k where r e,k =r e | m∈[1+(k-1)Q,kQ] ; In the connection of each sub-segment, a prediction mechanism is introduced to calculate the time unit n after the end of sub-segment k-1. p The target movement trend within the subsegment k is the first n p A known trajectory within a time unit, Among them, l pre,k is the time unit n after the end of subsegment k-1 p The target movement trend within the target; and establish the second relationship The target position information and velocity information provided by the rough trajectory are introduced into the measurement equation. The target position information is g3(x k )=||l crs,k -l k ||2 According to the target speed information, the equation is established as Using the equations established above, the problem of solving the state parameters of the moving target is transformed into a least squares problem The Gauss-Newton method is used to solve the above least squares problem and obtain the state of the target in each trajectory segment.

10. A wide-angle staring SAR moving target trajectory reconstruction device based on the Gauss-Newton method, characterized in that: include: The original echo processing module is used to perform sub-aperture division on the original radar echo signal to obtain a multi-channel original echo signal; Performing two-dimensional compression on the original echo signals of each channel in the range and azimuth dimensions to obtain the range-Doppler domain echo of each channel; performing preprocessing in the range-Doppler domain based on the range-Doppler domain echo of each channel, wherein the preprocessing includes multi-channel registration and clutter suppression to obtain the preprocessed range-Doppler domain echo; The parameter acquisition module extracts the phase history and envelope of the moving target based on the pre-processed range-Doppler domain echo and estimates the radial velocity of the moving target; The phase parameters of the moving target's phase history are coupled with the carrier position and the moving target's motion information. The maximum likelihood algorithm is used to estimate the phase parameters of each order in the extracted phase history. Based on the relationship between each order phase parameter and the motion parameters, the Gauss-Newton method is used to decouple the motion parameters and obtain the motion parameters of the moving target. The position reconstruction module obtains a rough motion trajectory reconstruction result based on the motion parameters of the moving target; the rough motion trajectory of the moving target is segmented, and the extracted envelope is decoupled from the rough motion trajectory of each segment using the Gauss-Newton method to obtain the precise reconstructed position of the moving target in each segment; The trajectory reconstruction module is used to splice the reconstructed position of the moving target in each segment of the motion trajectory to obtain a complete motion trajectory reconstruction result.

Images