Radar detection and HRRP imaging integrated method and system based on distance super-resolution

Abstract

The invention discloses a radar detection and HRRP imaging integrated method based on distance super-resolution, and belongs to the technical field of radar signal processing, and the method comprises the steps: S1, transmitting a narrowband LFM signal, receiving an echo, carrying out the pulse compression and coherent accumulation, and generating a distance-Doppler domain two-dimensional signal; s2, in a detection branch, extracting a target distance and a radial speed by adopting a traditional two-dimensional transverse false alarm rate (CFAR); s3, in an imaging branch, extracting a one-dimensional distance profile of a Doppler gate where a target is located, and performing super-resolution reconstruction through a regularization deconvolution algorithm; s4, fusing detection and imaging results, and outputting target parameters and a high-resolution range profile; according to the method, the unified LFM waveform is adopted, and the problem of time resource fragmentation caused by narrow / broadband signal switching is avoided.

Description

Technical Field

[0001] The present invention relates to the field of radar signal processing technology, and more particularly to a method for synchronously achieving target detection and high-resolution one-dimensional range profile (HRRP) in a narrowband linear frequency modulation (LFM) signal system. The method is particularly suitable for detecting and imaging moving targets by airborne radar. Background Art

[0002] High-resolution range profile (HRRP) imaging technology uses broadband linear frequency modulation (LFM) signals and pulse compression techniques to generate high-resolution scattering intensity distributions along the target's range dimension. Its key advantage lies in its intuitive principle and the lack of complex imaging conditions. Furthermore, HRRP can be combined with deep learning frameworks to achieve efficient matching and analysis by correlating the target's scattering characteristics with its motion information, providing crucial support for high-precision target recognition and classification.

[0003] Traditional HRRP imaging methods rely on prior target information: first, the radar system needs to obtain preliminary target position and distance information through conventional detection tasks, and then the dispatch center allocates wave positions according to the target position in the next mission cycle to perform HRRP imaging. This process separates target detection and HRRP imaging into different independent tasks, such as Figure 1 Specifically, radar detection uses narrowband linear frequency modulation (LFM) signals to match the size of conventional targets to maximize the signal-to-noise ratio. This is combined with Moving Target Detection (MTD) and Constant False Alarm Rate (CFAR) algorithms to achieve target detection and coarse position and velocity measurements. HRRP imaging requires switching to broadband LFM signals to improve range resolution and obtain fine-scale target scattering characteristics. This asynchronous task presents three problems: 1) Reduced real-time performance and accuracy. HRRP imaging results are highly sensitive to target pose. Specifically, target recognition requires real-time matching of the target's motion trajectory (output by the detection task) with the high-resolution range image (generated by the HRRP task). However, the time-sharing strategy results in processing delays in the sequential process of target detection, imaging, and recognition. Furthermore, for highly maneuverable targets, this delay significantly reduces the accuracy of matching pose and scattering signatures, directly impacting the accuracy of recognition results.

[0004] 2) Signal-to-noise ratio loss. To maintain the same search data rate, the time-sharing task mechanism compresses the pulse accumulation time during the detection and imaging phases, resulting in a decrease in signal processing gain. This severely degrades the target signal-to-noise ratio, especially at long distances or in low scattering intensity scenarios.

[0005] 3) High hardware complexity. Wideband imaging requires high sampling rate and large storage resources, which increases the system load and is not suitable for cost-constrained miniaturized unmanned airborne radars. SUMMARY

[0006] In view of the above problems, the purpose of the present application is to provide a target detection and HRRP imaging integrated method without signal bandwidth switching, to solve the problems of pulse resource waste, insufficient real-time performance and limited resolution caused by the fragmentation of the target detection and HRRP imaging process of the traditional radar system.

[0007] To achieve the above purpose, the technical scheme adopted by the present application is: a radar detection and HRRP imaging integrated method based on distance super-resolution, characterized by comprising the following steps: Step S1: transmitting a narrowband LFM signal, receiving a return signal and performing pulse compression and coherent accumulation to produce a two-dimensional range-Doppler domain signal; Step S2: in the detection branch, using a traditional two-dimensional cross false alarm rate CFAR to extract the target range and radial velocity; Step S3: in the imaging branch, extracting a one-dimensional range profile of the Doppler gate where the target is located, and performing super-resolution reconstruction through a regularization deconvolution algorithm; Step S4: fusing the detection and imaging results to output target parameters and a high-resolution range image.

[0008] Further, the step S3 specifically comprises: Step S31: extracting a one-dimensional range profile of the Doppler gate where the target is located, constructing a distance dimension super-resolution model to represent the coupling relationship between the target scattering distribution and the system response, and the expression is:

[0009] wherein, is the distance dimension observation data of the Doppler gate where the target is located after pulse compression and coherent accumulation processing of the radar multi-pulse original return signal, is a distance convolution kernel function, is a target scattering coefficient distribution, is a system receiver random noise.

[0010] Step S32: designing a sparse-smooth regularization objective function to simultaneously constrain the positioning of discrete strong scattering points and the recovery of continuous scattering structures; Step S33: solving the objective function based on iterative multiplication optimization to output a super-resolution HRRP image.

[0011] Further, the expression of the sparse-smooth regularization objective function is:

[0012] in, It is the range dimension observation data of the Doppler gate where the target is located after the radar multi-pulse original echo is processed by pulse compression and coherent accumulation. is the distance convolution kernel function, is the target scattering coefficient distribution, and Is the regularization parameter, which is used to adjust the weight ratio between the regularization term and the fidelity term, and is determined by the L curve method. is the optimal estimate of the target scattering coefficient distribution.

[0013] The iterative multiplication optimization adopts the Picard multiplication update rule, and the iterative formula is:

[0014] in, and are adjacent iterations, It is a solution A very small constant introduced by the norm non-differentiable problem, is a diagonal matrix, is a vector No. elements.

[0015] The present invention also discloses an integrated system of radar detection and HRRP imaging based on range super-resolution, which includes a narrowband LFM signal transceiver module, a range-Doppler domain processing module, a CFAR detection module and a regularized super-resolution imaging module.

[0016] Compared with the prior art, the technical solution adopted by the present invention has the following beneficial effects: (1) Using a unified LFM waveform avoids the time resource fragmentation problem caused by narrow / broadband signal switching; (2) The target detection information and HRRP image are output synchronously through a single processing flow, which significantly improves the real-time performance of the system and avoids the degradation of signal-to-noise ratio and accuracy caused by time segmentation; (3) It retains the low complexity advantage of narrowband systems and reduces hardware implementation costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the separation process of traditional target detection and HRRP imaging.

[0018] Figure 2 Schematic diagram of the flow of the integrated target detection and HRRP imaging method of the present invention.

[0019] Figure 3 Schematic diagram of the coupling relationship between target scattering distribution and system response.

[0020] Figure 4 Schematic diagram of target scattering coefficient distribution in the embodiment.

[0021] Figure 5 Schematic diagram of the echo (three-dimensional space) result after pulse compression in the embodiment.

[0022] Figure 6 Schematic diagram of the echo (planar projection) result after pulse compression in the embodiment.

[0023] Figure 7 Schematic diagram of echo (3D space) and target detection results after coherent accumulation in an embodiment.

[0024] Figure 8 Schematic diagram of the echo (planar projection) and target detection results after coherent accumulation in the embodiment.

[0025] Figure 9 for Figure 8 Schematic diagram of the range image of the Doppler gate where the target is located.

[0026] Figure 10 for Figure 9 Schematic diagram of the range image after super-resolution using this embodiment.

[0027] Figure 11 Schematic diagram of the target detection and HRRP imaging integrated system structure of the present invention. DETAILED DESCRIPTION

[0028] 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 them. Any modifications made based on the technical solutions in accordance with the technical ideas proposed by the present invention shall fall within the scope of protection of the present invention.

[0029] like Figure 2 As shown, this embodiment provides a method for integrating radar detection and HRRP imaging based on range super-resolution, including the following steps: Step S1: Transmit a narrowband LFM signal, receive the echo and perform pulse compression and coherent accumulation to produce a two-dimensional signal in the range-Doppler domain; Step S2: In the detection branch, the traditional two-dimensional cross-sectional false alarm rate (CFAR) is used to extract the target distance and radial velocity; Step S3: In the imaging branch, extract the one-dimensional range profile of the Doppler gate where the target is located, and perform super-resolution reconstruction using a regularized deconvolution algorithm; Step S31: Construct a range-Doppler domain echo convolution model to characterize the coupling relationship between the target scattering distribution and the system response, such as Figure 3 As shown; Step S32: design sparse-smooth regularization objective function to constrain discrete strong scattering point localization and continuous scattering structure recovery simultaneously; Step S33: solve the objective function based on iterative multiplication optimization, and output super-resolution HRRP image; Step S4: fuse detection and imaging results, and output target parameters and high-resolution range profile.

[0030] The expression of the sparse-smooth regularization objective function is as follows:

[0031] wherein, is the range dimension observation data of the Doppler gate where the target is located after pulse compression and coherent accumulation processing of radar multi-pulse original echo, is a range convolution kernel function, is a target scattering coefficient distribution, and is a weight parameter, is an optimal estimation value of the target scattering coefficient distribution.

[0032] The iterative multiplication optimization adopts a Picard multiplication update rule, and the iteration formula is as follows:

[0033] wherein, and are adjacent iteration times.

[0034] Embodiment: detection and HRRP imaging of a sea surface ship target by an airborne radar.

[0035] 1. System parameters: a. Center frequency: 10 GHz, LFM signal bandwidth: 20 MHz; pulse width: 20 us, pulse repetition frequency: 2000 Hz; b. Target distance: 3000 m, radial velocity: 6 m / s, size: 65 m, target signal-to-noise ratio: 30 dB, and the target range dimension scattering coefficient distribution is as shown in Figure 4 , which contains two groups of discrete impact-like scattering points and two groups of continuously distributed trapezoidal and rectangular scattering structures, respectively simulating the strong scattering point and the continuous scattering point in an actual target.

[0036] 2. Processing flow: a. The radar transmits a narrowband LFM signal, and receives 128 pulse echoes; b. Pulse compression processing is performed on the original echo, as shown in Figure 5 and Figure 6 ; and c. After pulse compression, the target energy is gathered by coherent accumulation processing, and the target detection is completed, the target distance is 3000m, and the radial velocity is 6m / s, as shown in Figure 7 and Figure 8 .

[0037] d. Extract and extract the one-dimensional range profile of the target Doppler gate (corresponding to the speed of 6m / s), as shown in Figure 9 . It can be seen that although the signal-to-noise ratio reaches 30dB, the distance resolution is insufficient due to the limitation of 20MHz signal bandwidth, resulting in scattering point aliasing, and a clear HRRP image cannot be obtained.

[0038] e. Input the regularization deconvolution target function, and determine the regularization weight parameter by using the L curve method; f. The optimal solution is obtained by using Picard multiplication method, and the super-resolution HRRP image is output, as shown in Figure 10 . It can be seen that compared with Figure 9 , the method of the present application successfully separates four groups of scattering points, which is close to the original scattering point distribution, and the structural similarity SSIM reaches 0.9433.

[0039] As shown in Figure 11 , the embodiment also discloses a radar detection and HRRP imaging integrated system based on distance super-resolution, which comprises a narrowband LFM signal transceiver module, a range-Doppler domain processing module, a CFAR detection module and a regularization super-resolution imaging module.

[0040] Although the present application has been disclosed as above with preferred embodiments, the embodiments and drawings are not intended to limit the present application, and any person skilled in the art can make various changes or modifications without departing from the spirit and scope of the present application, but also within the protection scope of the present application. Therefore, the protection scope of the present application should be limited by the protection scope of the claims of the present application.

Claims

1. A method for integrating radar detection and HRRP imaging based on range super-resolution, characterized in that: The following steps are involved: Step S1: Transmit a narrowband LFM signal, receive the echo and perform pulse compression and coherent accumulation to produce a two-dimensional signal in the range-Doppler domain; Step S2: In the detection branch, the traditional two-dimensional cross-sectional false alarm rate (CFAR) is used to extract the target distance and radial velocity; Step S3: In the imaging branch, extract the one-dimensional range profile of the Doppler gate where the target is located, and perform super-resolution reconstruction using a regularized deconvolution algorithm; Step S4: Fuse the detection and imaging results, and output target parameters and high-resolution range images.

2. The method for integrating radar detection and HRRP imaging based on range super-resolution according to claim 1, characterized in that: The step S3 specifically includes: Step S31: extract the one-dimensional range profile of the Doppler gate where the target is located, build a range-dimensional super-resolution model, and characterize the coupling relationship between the target scattering distribution and the system response. The expression is: ， in, It is the range dimension observation data of the Doppler gate where the target is located after the radar multi-pulse original echo is processed by pulse compression and coherent accumulation. is the distance convolution kernel function, is the target scattering coefficient distribution, is the random noise of the system receiver, Step S32: designing a sparse-smooth regularization objective function to simultaneously constrain the positioning of discrete strong scattering points and the recovery of continuous scattering structures; Step S33: Solve the objective function based on iterative multiplication optimization and output a super-resolution HRRP image.

3. The method for integrating radar detection and HRRP imaging based on range super-resolution according to claim 2, characterized in that: The expression of the sparse-smooth regularization objective function is: ， in, It is the range dimension observation data of the Doppler gate where the target is located after the radar multi-pulse original echo is processed by pulse compression and coherent accumulation. is the distance convolution kernel function, is the target scattering coefficient distribution, and Is the regularization parameter, which is used to adjust the weight ratio between the regularization term and the fidelity term, and is determined by the L curve method. is the optimal estimate of the target scattering coefficient distribution.

4. The method for integrating radar detection and HRRP imaging based on range super-resolution according to claim 2, characterized in that: The iterative multiplication optimization adopts the Picard multiplication update rule, and the iterative formula is: ， in, and is the number of adjacent iterations, It is a solution A very small constant introduced by the norm non-differentiable problem, is a diagonal matrix, is a vector No. elements.

5. A system for implementing the method according to any one of claims 1 to 4, characterized in that: include: Narrowband LFM signal transceiver module, pulse compression module, range-Doppler domain processing module, CFAR detection module and regularized super-resolution imaging module; The narrowband LFM signal transceiver module is responsible for generating narrowband LFM pulse train signals, down-sampling them, and then transmitting the signals to the pulse compression module; The pulse compression module performs pulse compression processing on the received narrowband LFM pulse signals one by one and sends the processing results to the range-Doppler processing module; The range-Doppler processing module performs coherent accumulation on the pulse train data after pulse compression along the slow time dimension to extract Doppler information, and transmits the processing results to the CFAR detection module and the regularized super-resolution imaging module simultaneously; The CFAR detection module performs constant false alarm rate (CFAR) target detection on the data from the range-Doppler processing module and outputs target detection information. The regularized super-resolution imaging module extracts the one-dimensional distance data of each Doppler unit in turn, and generates high-resolution range profile (HRRP) information through regularized super-resolution imaging processing.

Images