Forward-looking super-resolution imaging method and device for high-speed platform-mounted small array radar

Abstract

The invention discloses a high-speed platform-borne small array radar foresight super-resolution imaging method and device, and belongs to the technical field of digital signal processing, and the method comprises the steps: obtaining an echo signal generated by a strong scattering point in a radar foresight super-resolution imaging scene, and carrying out the carrier frequency removal processing; sequentially performing pulse compression, range walk correction and Doppler frequency offset correction processing on the obtained echo signal subjected to carrier frequency removal to obtain a corrected echo signal; constructing a foresight super-resolution imaging model according to the corrected echo signal; performing sparse reconstruction on the foresight super-resolution imaging model by using an imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image; and correcting the super-resolution image according to a pre-calculated space-variant phase error to obtain a corrected super-resolution image. According to the method, image distortion caused by azimuth space-variant errors can be effectively corrected, the focusing effect and precision of the corrected super-resolution image are improved, and the method is suitable for practical application scenes.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of digital signal processing, and particularly relates to a high-speed platform-mounted small array radar forward-looking super-resolution imaging method and device. BACKGROUND

[0002] In recent years, the development of radar technology makes radar imaging more and more important. In actual application scenarios, high supersonic aircraft has become a research hotspot. Such a platform often has a flight speed of several times or even dozens of times the speed of sound. How to realize radar high-resolution forward-looking imaging on a high-speed moving platform is of great significance. Due to the particularity of such a platform, the imaging system has high requirements on system structure, imaging real-time performance, and imaging resolution.

[0003] The forward-looking imaging scene generally refers to the range of -10°~10° in front of the radar. In this scene, the Doppler frequency gradient of the echo signal is small. The traditional synthetic aperture radar (SAR) imaging technology is limited to the forward-looking scene. The improved range-Doppler (RD) algorithm and the improved line frequency modulation scale (Chrip Scaling, CS) algorithm for time domain correction of range walk can only achieve an imaging squint angle of about 70°, which cannot be applied to the forward-looking imaging scene. In order to realize forward-looking high-resolution imaging, researchers have proposed a scanning beam imaging technology based on a real aperture. This technology processes the azimuth echo obtained after radar scanning to obtain a forward-looking scene image. However, in actual scenarios, high-speed platforms are mostly small in size, and the aperture of the radar carried by the platform is much smaller than that of airborne radar and shipborne radar. Therefore, the azimuth resolution is poor, and it is difficult to meet the requirements of radar forward-looking high-resolution imaging. In recent years, super-resolution technology has attracted widespread attention due to its ability to break through the resolution limit and simple implementation.

[0004] However, the existing method using super-resolution technology has problems such as platform motion speed much smaller than the speed of sound and imaging result error, and therefore is not suitable for the case where the platform motion speed is several times the speed of sound. SUMMARY

[0005] In order to solve the above problems in the prior art, the application provides a high-speed platform-mounted small array radar forward-looking super-resolution imaging method and device.

[0006] The technical problem to be solved by the application is solved by the following technical scheme: In a first aspect, the application provides a high-speed platform-mounted small array radar forward-looking super-resolution imaging method, which comprises: Obtain echo signals generated by a strong scattering point in a radar forward-looking super-resolution imaging scene, and perform de-carrier frequency processing on the echo signals to obtain the echo signals after de-carrier frequency processing; Perform pulse compression, range walk correction and Doppler frequency offset correction processing on the echo signals after de-carrier frequency processing in sequence to obtain the echo signals after correction; Construct a forward-looking super-resolution imaging model according to the echo signals after correction; Perform sparse reconstruction on the forward-looking super-resolution imaging model by using an imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image; Correct the super-resolution image according to a pre-computed space-varying phase error to obtain the super-resolution image after correction.

[0007] Optionally, the performing pulse compression, range walk correction and Doppler frequency offset correction processing on the echo signals after de-carrier frequency processing in sequence to obtain the echo signals after correction comprises: Perform pulse compression on the echo signals after de-carrier frequency processing to obtain a signal after pulse compression; Perform FFT on the signal after pulse compression along a fast time dimension to obtain a transformed signal; Multiply the transformed signal by a pre-constructed linear frequency phase factor and then perform IFFT to obtain a signal after range walk correction; Obtain a signal after Doppler frequency offset correction according to the signal after range walk correction and a pre-constructed Doppler frequency offset correction factor; Correct the signal after Doppler frequency offset correction according to a pre-computed space-varying phase error to obtain the echo signals after correction.

[0008] Optionally, the pre-constructed linear frequency phase factor is represented as follows: ; wherein, denotes an exponential function, denotes an imaginary unit, denotes a speed of light, denotes a distance coordinate of the strong scattering point in a plane of a rectangular coordinate system with a radar position as an origin, denotes a radar platform motion speed, denotes a distance coordinate, denotes a slow time, denotes a distance frequency; The pre-constructed Doppler frequency offset correction factor is represented as follows: ; wherein, denotes a signal wavelength.

[0009] Optionally, the space-varying phase error is represented as follows: ; wherein, represents the space-varying phase error, represents the azimuth coordinate of the strong scattering point in the plane of the rectangular coordinate system with the radar position as the origin, represents the central slant range.

[0010] Optionally, the constructing a forward-looking super-resolution imaging model according to the corrected echo signal comprises: taking out any one range cell data from the corrected echo signal; constructing a dictionary matrix according to a predefined discrete Doppler sequence; constructing the forward-looking super-resolution imaging model according to the dictionary matrix.

[0011] Optionally, the correcting the super-resolution image according to the pre-computed space-varying phase error to obtain a corrected super-resolution image comprises: dividing a preset image into a grid using the azimuth resolution and the range resolution of the super-resolution image to obtain a divided preset image; wherein the super-resolution image and the preset image have the same size; calculating the coordinates of each pixel point in the super-resolution image at the corresponding pixel points in the divided preset image according to the space-varying phase error to obtain corrected coordinates; performing geometric correction on the divided preset image based on sinc kernel interpolation according to the corrected coordinates to obtain the corrected super-resolution image.

[0012] Optionally, the corrected coordinates are represented as follows: ; wherein, represents the coordinates of the pixel point in the super-resolution image, represents the coordinates of the pixel point in the divided preset image corresponding to the pixel point at the coordinates in the super-resolution image, represents the space-varying phase error, represents the signal wavelength, represents the height of the radar platform, represents the movement speed of the radar platform, represents the azimuth equivalent movement speed of the array element.

[0013] In a second aspect, the present application provides a forward-looking super-resolution imaging device for a high-speed platform-mounted small array radar, which comprises: ​The signal acquisition module is configured to acquire echo signals generated by strong scattering points in a radar forward-looking super-resolution imaging scene, and to perform de-carrier frequency processing on the echo signals to obtain the echo signals after de-carrier frequency processing. The signal correction module is configured to sequentially perform pulse compression and range walk correction processing on the echo signals after de-carrier frequency processing to obtain the echo signals after correction. The model construction module is configured to construct a forward-looking super-resolution imaging model according to the echo signals after correction. The image acquisition module is configured to perform sparse reconstruction on the forward-looking super-resolution imaging model by using an imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image. The image correction module is configured to correct the super-resolution image according to a pre-computed space-variant phase error to obtain the super-resolution image after correction.

[0014] The technical scheme provided by the embodiment of the present application can include the following beneficial effects: In the above technical scheme, the present application considers the influence of azimuth space-variant error on the signals after range walk correction, and has higher precision; the present application uses an imaging algorithm based on sparse Bayesian learning to perform sparse reconstruction on the forward-looking super-resolution imaging model, the algorithm can automatically estimate super parameters in the iteration process, and is more suitable for actual application scenarios; and the present application performs geometric correction on the super-resolution image according to the space-variant phase error, can effectively correct the image distortion caused by the azimuth space-variant error, and the super-resolution image after correction has better focusing effect.

[0015] The present application will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1 is a flowchart of a high-speed platform-mounted small array radar forward-looking super-resolution imaging method provided by an embodiment of the present application; Figure 2 is a signal propagation schematic diagram of a high-speed motion platform array in a radar single-transmit-single-receive mode provided by an embodiment of the present application; Figure 3 is a scene schematic diagram provided by an embodiment of the present application; Figure 4 is a real beam imaging result schematic diagram provided by an embodiment of the present application; Figure 5 is a result schematic diagram of a traditional super-resolution imaging algorithm; Figure 6 is an imaging result schematic diagram provided by an embodiment of the present application; Figure 7 is a strong scattering point schematic diagram provided by an embodiment of the present application; Figure 8is another imaging result schematic view provided by an embodiment of the present application; Figure 9 is a result schematic view of a traditional super-resolution imaging algorithm in another scenario; Figure 10 is another imaging result schematic view provided by an embodiment of the present application; Figure 11 is a block diagram of a high-speed platform-mounted small array radar forward-looking super-resolution imaging device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0017] The present application will be further described in detail below in conjunction with specific embodiments, but the embodiments of the present application are not limited thereto.

[0018] Figure 1 is a flowchart of a high-speed platform-mounted small array radar forward-looking super-resolution imaging method provided by an embodiment of the present application, as shown in the figure, the method can include the following steps: Figure 1 S101, obtaining echo signals generated by a strong scattering point in a radar forward-looking super-resolution imaging scene, and performing de-carrier frequency processing on the echo signals to obtain de-carrier frequency echo signals.

[0019] In an embodiment, Figure 2 is a signal propagation schematic view of a high-speed motion platform array in a radar single-transmit-single-receive mode provided by an embodiment of the present application, a forward-looking super-resolution imaging rectangular coordinate system plane can be established with the radar position as the origin, and a strong scattering point P in the imaging scene is taken as an example , wherein, represents the range direction coordinate of the strong scattering point P in the rectangular coordinate system plane, represents the azimuth direction coordinate of the strong scattering point P in the rectangular coordinate system plane, and the LFM (Linear Frequency Modulation) signal transmitted by the array element in turn when the radar is working can be represented as follows: ; , wherein, represents the LFM signal, represents the fast time, represents the range window function, is the carrier frequency, is the modulation slope, is the pulse width, represents the rectangular window function, represents the exponential function, represents the imaginary unit.

[0020] ​According to the LFM signal transmitted by the array elements in turn when the radar works, the echo signal generated by the strong scattering point P, after the de-carrier frequency processing, the de-carrier frequency echo signal obtained can be expressed as follows: ; Wherein, represents the de-carrier frequency echo signal, represents the total time of the array element transmitting signal, is the signal amplitude, represents the slow time, represents the position of the array element from the strong scattering point P at the slow time , is the speed of light, is the signal wavelength.

[0021] The is carried out second-order Taylor expansion: ; Wherein, the central slant distance , represents the height of the radar platform, represents the speed of the radar platform movement, represents the equivalent movement speed of the array element in the azimuth direction. In the forward-looking imaging scene is much larger than , and is much larger than and . Therefore, the phase difference caused by the quadratic term is much smaller than , so the quadratic term about can be discarded, and the following is obtained: .

[0022] S102, the de-carrier frequency echo signal is sequentially processed by pulse compression, range walk correction and Doppler frequency offset correction to obtain the corrected echo signal.

[0023] Optionally, S102 can include: pulse compression is performed on the de-carrier frequency echo signal to obtain a pulse compressed signal; performing FFT (Fast Fourier Transform) on the pulse compressed signal along the fast time dimension to obtain a transformed signal; multiplying the transformed signal and the pre-constructed linear frequency phase factor and then performing IFFT (Inverse Fast Fourier Transform) to obtain a range walk corrected signal; ​The Doppler frequency offset corrected signal is obtained by using the distance-corrected signal and the pre-constructed Doppler frequency offset correction factor; The Doppler frequency offset corrected signal is corrected based on the spatially varying phase error to obtain the corrected echo signal.

[0024] The echo signal after carrier frequency removal is pulse compressed to obtain the pulse-compressed signal: ; in, Indicates signal bandwidth. Let Singer's function be represented. A Fast Fourier Transform (FFT) is performed on the pulse-compressed signal along the fast time dimension, resulting in the following transformed signal representation: ; in, For range frequency, according to It can be known that the increase in distance traveled by the echo per unit time point is ,because Unknown, therefore can be adopted To approximate ,here, Represents the range coordinates. A linear frequency phase factor is pre-constructed as follows: After multiplying with the transformed signal, perform an IFFT to obtain the distance-corrected signal: ; Will Substituting the distance-corrected signal into the equation, we obtain a new expression for the distance-corrected signal: ; in, It can be considered a constant within a distance cell. The Doppler frequency is caused by the equivalent motion of the array antenna and contains the azimuth information of the target (strong scattering point P). The Doppler frequency is caused by the platform's motion.

[0025] because Unknown, therefore can be adopted To approximate The Doppler frequency offset correction factor is pre-constructed as follows: The Doppler frequency offset corrected signal, obtained from the distance-travel-corrected signal and the pre-constructed Doppler frequency offset correction factor, can be expressed as follows: ; The spatially variable phase error is expressed as follows: ; wherein, denotes the range-dependent phase error, denotes the range coordinate of the strong scatterer in the plane of the Cartesian coordinate system with the radar position as the origin, denotes the azimuth coordinate of the strong scatterer in the plane of the Cartesian coordinate system with the radar position as the origin, denotes the radar platform motion velocity, denotes the central slant range.

[0026] Substituting the range-dependent phase error into the Doppler frequency offset corrected signal, the corrected echo signal is obtained, denoted as follows: ; wherein, denotes the signal bandwidth, denotes the sinc function.

[0027] S103, constructing a forward-looking super-resolution imaging model according to the corrected echo signal.

[0028] Optionally, S103 can include: taking out any one range cell data from the corrected echo signal; constructing a dictionary matrix according to the predefined discrete Doppler sequence; constructing a forward-looking super-resolution imaging model according to the dictionary matrix.

[0029] Specifically, any one range cell data is taken out from the corrected echo signal, denoted as an observation signal , which can be expressed as: ; wherein, , denotes the Gaussian noise, denotes the amplitude of the signal at the th strong scatterer, denotes the number of strong scatterers in the range cell. The discrete Doppler sequence is defined as , wherein, is the Doppler frequency interval, is the pulse repetition frequency, , the dictionary matrix is constructed , denoted as: ; wherein, denotes the th element in the slow time.

[0030] Then, the range cell data, i.e., the observation signal Rewritten as the product of the Fourier matrix and the sparse signal, it can be expressed as follows: ; where the sparse signal contains the position and amplitude information of the strong scattering points.

[0031] S104, sparse reconstruction is performed on the forward-looking super-resolution imaging model by using the imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image.

[0032] Specifically, the process of sparse reconstruction of the forward-looking super-resolution imaging model by using the imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image can be referred to as follows: Initialize the hyperparameters of the sparse Bayesian learning algorithm and wherein is used to describe the state of each element in , and represents the variance of the Gaussian noise. Let , , and the residual , represents the i-th element in , the input observation signal , and the dictionary matrix , and initialize , , , the length of , , the length of .

[0033] Assuming that each element in the recovered signal is subject to a normal distribution with a mean of 0 and a variance of , then The probability density function of can be expressed as: ; wherein represents the i-th element in , , and represents the probability density.

[0034] According to , the Gaussian likelihood model under the sparse Bayesian learning algorithm framework can be expressed as: ; By using the Bayesian formula, the edge probability density function of the observation signal about can be obtained as: ; in, variance diagonal matrix , The identity matrix is ​​the sparse signal to be estimated. Given the observed signal Regarding the situation The marginal probability density function is: ; in, The above formula also follows a Gaussian distribution, so under the current parameters... The maximum a posteriori (MAP) estimate can be expressed as: ; ; in, Marginal probability density function The mean, Marginal probability density function The variance.

[0035] This invention employs the Expectation Maximization (EM) algorithm to estimate hyperparameters and cost functions. It can be represented as: ; Estimation under the old hyperparameter conditions Cost function The first item and It is irrelevant; in the E-step of the EM algorithm (i.e., calculating the cost function), the cost function can be... Simplified to cost function : ; in, It indicates a desire for the expected value.

[0036] because , This indicates the calculation of variance. Represents the first sparse signal Substitute each element into the cost function. have to: ; in, Marginal probability density function The first of the mean One element; In the M-step (maximizing the cost function) of the EM algorithm, the cost function is maximized. To obtain the parameters: ; right Take the derivative, set it to 0, and proceed to the hyperparameter. Update formula: ; Fixed update ,renew At this point, the cost function The second item in Regardless, in step E, the cost function Simplified to : ; because, , Represents any matrix, The mean, For variance, This indicates the transpose operation, for Simplify the second term: ; in, for Expectations The trace of the matrix; From the marginal probability density function available: ; Substituting the above equation into... ,get: ; in, These are the hyperparameters from the previous iteration. In the M-step, for... By taking the derivative and setting it to zero, we obtain the hyperparameters. Update formula: ; Determine the sparse signal under the current conditions If convergence is detected, or if the maximum number of iterations is reached, the loop exits and a sparse signal is output. and for sparse signals The absolute value of each element is taken to obtain the super-resolution image. Then, the next range data is processed. If the condition is not met, the calculation is returned. about probability density function The steps.

[0037] The convergence condition can be expressed as: ; in, This indicates the result of the previous iteration.

[0038] S105. Correct the super-resolution image based on the pre-calculated spatial phase error to obtain the corrected super-resolution image.

[0039] Optionally, S105 may include: The preset image is divided into grids using the azimuth and range resolutions of the super-resolution image to obtain the divided preset image; wherein, the super-resolution image and the preset image have the same size; Based on the spatially variable phase error, the coordinates of each pixel in the super-resolution image are calculated to the corresponding pixel in the pre-defined image after segmentation, and the corrected coordinates are obtained. Based on the corrected coordinates, the pre-defined image after segmentation is geometrically corrected using sinc kernel interpolation to obtain the corrected super-resolution image.

[0040] Understandably, using super-resolution images The azimuth resolution and range resolution of the preset image The image is divided into a uniform grid, with each pixel representing a small region of the imaging scene. This is useful for super-resolution images. and preset images In the preset image In the middle The pixel at a given location can be used to calculate its corresponding super-resolution image based on the spatially varying phase error. pixels in coordinate: The corrected coordinates are represented as follows: ; in, This represents the coordinates of a pixel in the super-resolution image. This indicates the relationship between the predefined image after segmentation and the super-resolution image. The coordinates of the pixel corresponding to the pixel at that location. This indicates the spatially variable phase error. Indicates the signal wavelength. Indicates the altitude of the radar platform. Indicates the speed of the radar platform. It represents the equivalent velocity of the array element's azimuth direction.

[0041] The detailed derivation process can be found below: When the platform is moving at low speed or stationary, it is in a distance-oriented position. Orientation Doppler frequency of echo at the target point is: ; It should be noted here that , wherein is the range coordinate of the target point on the plane, is the azimuth coordinate of the target point on the plane.

[0042] When the platform motion speed is , then the echo Doppler frequency becomes: ; Because , , it can be simplified as: ; The final coordinate conversion formula is: .

[0043] From the super-resolution image , find the point closest to , and take the data of the area centered on the point as . The corresponding sinc interpolation kernel is: ; wherein , , denotes the range resolution, denotes the azimuth resolution after super-resolution, and the value at is: ; wherein denotes the Hadamard product of the matrix, and after traversing all pixel points, the corrected super-resolution image is obtained.

[0044] The effect of the present application is further illustrated by the following simulation experiment: Simulation conditions: The parameters of the high-speed moving platform small array radar system are set, as shown in Table 1.

[0045] Table 1

[0046] In the simulation experiment, the traditional super-resolution algorithm is compared with the present application. ​

[0047] Simulation content: Experiment 1. Three point targets are set in the range direction 6km, at-57m, 0m, 45m respectively, Figure 3 is a scene schematic diagram provided by an embodiment of the application, Figure 4 is a real beam imaging result schematic diagram provided by an embodiment of the application, Figure 5 is a result schematic diagram of a traditional super-resolution imaging algorithm, Figure 6 is an imaging result schematic diagram provided by an embodiment of the application. Figures 3-6 are all normalized, Figures 3-6 It is shown that the real beam imaging result cannot distinguish the three points, the traditional super-resolution imaging algorithm result appears offset, the target points appear at-70m, 0m, 35m respectively, and the two points on the left and right of the scene center deviate from the ideal positions due to the azimuth direction variation error, the target points of the imaging result of the application appear at-58.5m, 0m, 46.8m respectively, which conforms to the pre-set scene.

[0048] Experiment 2. An imaging scene of a ship target is set, and one interference source is placed on the left and right of the ship, Figure 7 is a strong scattering point schematic diagram provided by an embodiment of the application, Figure 8 is a real beam imaging result schematic diagram provided by an embodiment of the application, Figure 9 is a result schematic diagram of a traditional super-resolution imaging algorithm in another scene, Figure 10 is another imaging result schematic diagram provided by an embodiment of the application. Figures 7-10 It is shown that the real beam imaging result cannot distinguish the ship target and the interference sources on the left and right, the traditional super-resolution imaging algorithm result can distinguish the ship target and the interference sources, but the position of the recovered strong scattering point deviates from the original position, the imaging result of the application is more accurate, eliminates image offset, and accurately reconstructs the scene.

[0049] Figure 11 is a block diagram of a high-speed platform carrying a small array radar forward-looking super-resolution imaging device provided by an embodiment of the application, as shown in Figure 11 The device 1100 can include: A signal acquisition module 1101, configured to acquire echo signals generated by strong scattering points in a radar forward-looking super-resolution imaging scene, and perform de-carrier frequency processing on the echo signals to obtain echo signals after de-carrier frequency processing; A signal correction module 1102, configured to sequentially perform pulse compression and range walk correction processing on the echo signals after de-carrier frequency processing to obtain corrected echo signals; A model construction module 1103, configured to construct a forward-looking super-resolution imaging model according to the corrected echo signals; The image acquisition module 1104 is configured to perform sparse reconstruction on the forward-looking super-resolution imaging model by using the imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image. The image correction module 1105 is configured to correct the super-resolution image according to the pre-calculated spatially variant phase error to obtain a corrected super-resolution image.

[0050] The application considers the influence of the azimuthally variant error on the distance migration corrected signal, and has higher accuracy; the imaging algorithm based on sparse Bayesian learning is used to perform sparse reconstruction on the forward-looking super-resolution imaging model, the algorithm can automatically estimate the hyperparameters in the iteration process, and is more suitable for actual application scenarios; and the super-resolution image is geometrically corrected according to the spatially variant phase error, the image distortion caused by the azimuthally variant error can be effectively corrected, and the focusing effect of the corrected super-resolution image is better.

[0051] It should be noted that the terms "first", "second", and the like are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of devices and methods consistent with some aspects of the present application.

[0052] In the description of the present application, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" means that the specific features or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present application, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features 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 different embodiments or examples described in the present application.

[0053] Although the present application is described herein in conjunction with various embodiments, other variations of the disclosed embodiments can be understood and implemented by those skilled in the art by viewing the drawings and the disclosure. In the description of the present application, the word "comprising" does not exclude other components or steps, "one" or "an" does not exclude a plurality, and "plurality" means two or more, unless otherwise explicitly specified. In addition, some measures are described in different embodiments, but this does not mean that these measures cannot be combined to produce good results.

[0054] The above is further detailed description of the present application in combination with specific preferred embodiments, and cannot be deemed as limitation of the specific implementation of the present application to these descriptions. For those skilled in the art to which the present application belongs, without departing from the concept of the present application, a number of simple deductions or substitutions can be made, and all should be deemed as falling within the protection scope of the present application.

Claims

1. A high-speed platform-borne small array radar forward-looking super-resolution imaging method, characterized in that, The method comprises: acquiring echo signals generated by strong scattering points in a radar forward-looking super-resolution imaging scene, and performing de-carrier frequency processing on the echo signals to obtain the echo signals after de-carrier frequency processing; sequentially performing pulse compression, range walk correction and Doppler frequency offset correction processing on the echo signals after de-carrier frequency processing to obtain the echo signals after correction; constructing a forward-looking super-resolution imaging model according to the echo signals after correction; performing sparse reconstruction on the forward-looking super-resolution imaging model by using an imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image; correcting the super-resolution image according to a pre-computed space-variant phase error to obtain the super-resolution image after correction.

2. The high-speed platform-borne small array radar forward-looking super-resolution imaging method according to claim 1, characterized in that, The sequentially performing pulse compression, range walk correction and Doppler frequency offset correction processing on the echo signals after de-carrier frequency processing to obtain the echo signals after correction comprises: performing pulse compression on the echo signals after de-carrier frequency processing to obtain signals after pulse compression; performing FFT on the signals after pulse compression along a fast time dimension to obtain transformed signals; multiplying the transformed signals and a pre-constructed linear frequency phase factor to perform IFFT to obtain signals after range walk correction; obtaining signals after Doppler frequency offset correction according to the signals after range walk correction and a pre-constructed Doppler frequency offset correction factor; correcting the signals after Doppler frequency offset correction according to a pre-computed space-variant phase error to obtain the echo signals after correction.

3. The high-speed platform borne small array radar forward looking super resolution imaging method of claim 2, wherein, The pre-constructed linear frequency phase factor is expressed as follows: ； wherein, denotes the exponential function, denotes the imaginary unit, denotes the speed of light, denotes the distance coordinate of the strong scattering point in the plane of the Cartesian coordinate system with the radar position as origin, denotes the radar platform movement velocity, denotes the distance coordinate, denotes the slow time, denotes the range frequency; The pre-constructed Doppler frequency offset correction factor is expressed as follows: ； wherein denotes the signal wavelength.

4. The high-speed platform-borne small array radar forward-looking super-resolution imaging method according to claim 3, characterized in that, The space-variant phase error is expressed as follows: ； wherein denotes the spatially variant phase error, denotes the azimuth coordinate of the strong scattering point in the plane of the Cartesian coordinate system with the radar position as origin, denotes the central slant range.

5. The high-speed platform-borne small array radar forward-looking super-resolution imaging method according to claim 1, characterized in that, The constructing a forward-looking super-resolution imaging model according to the echo signals after correction comprises: taking out any one distance unit data from the echo signals after correction; constructing a dictionary matrix according to a pre-defined discrete Doppler sequence; constructing the forward-looking super-resolution imaging model according to the dictionary matrix.

6. The high-speed platform-borne small array radar forward-looking super-resolution imaging method according to claim 1, characterized in that, The correcting the super-resolution image according to a pre-computed space-variant phase error to obtain the super-resolution image after correction comprises: dividing a preset image into grids by using the azimuth direction resolution and the range direction resolution of the super-resolution image to obtain the preset image after division; wherein the super-resolution image and the preset image have the same size; calculating coordinates of each pixel point in the super-resolution image at corresponding pixel points in the preset image after division according to the space-variant phase error to obtain corrected coordinates; performing geometric correction on the preset image after division based on sinc kernel interpolation according to the corrected coordinates to obtain the super-resolution image after correction.

7. The high-speed platform-borne small array radar forward-looking super-resolution imaging method according to claim 6, characterized in that, The corrected coordinates are expressed as follows: ； wherein, denotes a coordinate of a pixel point in the super-resolution image, denotes a coordinate of a pixel point in the divided preset image corresponding to the pixel point at denotes a coordinate of a pixel point in the divided preset image corresponding to the pixel point at denotes the spatially variant phase error, denotes a signal wavelength, denotes a height of a radar platform, denotes a velocity of a radar platform, denotes an equivalent velocity of an array element in an azimuth direction.

8. A high-speed platform-borne small array radar forward-looking super-resolution imaging device, characterized in that The device comprises: a signal acquisition module configured to acquire echo signals generated by strong scattering points in a radar forward-looking super-resolution imaging scene, and perform de-carrier frequency processing on the echo signals to obtain the echo signals after de-carrier frequency processing; a signal correction module configured to sequentially perform pulse compression and range walk correction processing on the echo signals after de-carrier frequency processing to obtain the echo signals after correction; and a super-resolution image construction module configured to construct a forward-looking super-resolution imaging model according to the echo signals after correction, perform sparse reconstruction on the forward-looking super-resolution imaging model by using an imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image, and correct the super-resolution image according to a pre-computed space-variant phase error to obtain the super-resolution image after correction. a model construction module, configured to construct a forward-looking super-resolution imaging model according to the corrected echo signal; an image acquisition module, configured to perform sparse reconstruction on the forward-looking super-resolution imaging model by using an imaging algorithm based on sparse Bayesian learning to obtain a super-resolution image; an image correction module, configured to correct the super-resolution image according to a pre-computed spatially variant phase error to obtain a corrected super-resolution image.

Images