Navigation radar beam sharpening method based on peak valley detection and matching pursuit

Abstract

The invention discloses a navigation radar wave beam sharpening method based on peak valley detection and matching pursuit, and the method comprises the steps: supposing that two ships are located at the same detection distance of a navigation radar, but the azimuth angle difference is a small delta theta, and if the navigation radar just can judge that the two ships are two different targets, the two ships are located at the same detection distance; the method comprises the following steps of: firstly, taking the radar as a ship instead of taking the radar as a ship, namely, the azimuth resolution of the navigation radar is delta theta, and distinguishing targets of which the azimuth interval exceeds half-power beam width or peak valleys are formed on azimuth-dimension echoes due to a relatively large relative phase difference value by adopting a peak-valley detection method; and for a single-peak target formed on the azimuth-dimension echo, judging whether the target is formed by superposing two single-target echoes with the same phase or not through matching search. According to the method, the azimuth resolution of the navigation radar can break through the limitation of the antenna aperture, the resolution capability of the navigation radar on two targets with close azimuth above the same distance unit is improved, and the target positioning and tracking performance of the navigation radar is further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of radar signal processing, and in particular to a navigation radar beam sharpening method based on peak-valley detection and matching pursuit. Background Art

[0002] To accurately determine target positions and ensure good tracking performance during multi-target intersections and when targets approach the shore, navigation radars require high azimuth resolution. Navigation radars acquire target azimuth information using a single-antenna mechanical scanning system, and their azimuth resolution is limited by the antenna's azimuth beamwidth.

[0003] Due to the limitation of antenna size (the installation conditions of some small ships do not support it, and an overly large antenna will also bring additional pressure to the structure and servo), the azimuth beam width of the navigation radar antenna cannot be made very narrow, which is not conducive to distinguishing targets with close intervals (such as Figure 1 As shown in Figure 2, two maneuvering targets are also prone to trajectory exchange when they approach each other. Summary of the Invention

[0004] The present invention proposes a beam sharpening method based on peak-valley detection and matching search. By segmenting radar echo signals that exceed the detection threshold, the segmented signals are then subjected to peak-valley detection (each peak corresponds to a target). Matching search is then used to determine whether an echo signal segment containing only one peak can be decomposed into the superposition of two single-target echoes. This beam sharpening method can improve the radar's azimuth resolution without changing the antenna size, thereby greatly enhancing the navigation radar's target positioning, differentiation, and tracking capabilities.

[0005] The present invention is achieved through the following technical solutions:

[0006] A navigation radar beam sharpening method based on peak-valley detection and matching pursuit is proposed. For a navigation radar, assume that there are two ships at the same detection range of the radar, but with a small azimuth difference of Δθ. If the navigation radar can just identify the two ships as two different targets rather than treating them as a single ship, the azimuth resolution of the navigation radar is called Δθ. For targets whose azimuth interval exceeds the half-power beamwidth or whose azimuth echoes form peaks and valleys due to large relative phase differences, the peak-valley detection method is used to distinguish them. For targets with single peaks in the azimuth echo, a matching search is performed to determine whether they are the superposition of two single-target echoes with the same phase. The specific operation steps are as follows:

[0007] (1) Echo segmentation: First, perform echo segmentation on the azimuth echo to extract the continuous echo area above the detection threshold;

[0008] (2) Peak and valley detection: Then, a dynamic threshold generated based on the maximum value point is used to perform peak and valley detection on each continuous echo area;

[0009] (3) Matching search: Secondly, for the continuous echo area with only one peak, a matching search algorithm is used to further determine whether it is composed of the superposition of two single target echoes;

[0010] (4) Echo generation: Finally, the matching search results and the sharpened beam pattern are used to generate sharpened echoes, which are then fused with the peak and valley detection results to form azimuth-dimensional beam-sharpened echoes.

[0011] As a preferred embodiment, in step (1), the azimuth echo data is traversed according to the index. During this process, continuous areas that are higher than a preset detection threshold and whose length is greater than a specific value are extracted, and the starting positions of these continuous areas are stored.

[0012] As a preferred embodiment, in step (2), a dynamic threshold is used to perform peak and valley detection on each continuous area. The dynamic threshold is generated by generating a detection platform slightly lower than the amplitude value of the data point near the neighborhood of each data point. The maximum value of the overlapping detection platforms is taken as the dynamic threshold to avoid random noise interference. The area above the dynamic threshold is the peak area.

[0013] As a preferred embodiment, in step (3), for the case where there is only one peak in the continuous area, a specific matching search algorithm is used to perform beam sharpening processing, and the least squares method is used to determine whether the single-peak continuous area has the possibility of being decomposed into two superimposed single target echoes.

[0014] As a preferred embodiment, in step (4), if there are more than two peaks in the continuous area, the dynamic threshold detection result is used as the sharpened echo; if there is only one peak in the continuous area, the sharpened echo is generated using the antenna sharpening pattern based on the two single target positions and amplitudes obtained by the matching search algorithm.

[0015] As a preferred embodiment, the azimuth resolution of the navigation radar is approximately equal to the half-power beamwidth of the navigation radar antenna. The relative phase of two targets will affect the target resolution process, and it is most difficult to resolve when the two targets have the same phase.

[0016] Beneficial effects: The present invention can enable the azimuth resolution of the navigation radar to break through the limitation of the antenna aperture, improve the navigation radar's ability to resolve two targets with close azimuths in the same distance unit, and thus improve the navigation radar's positioning and tracking performance of the target. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of sea surface echo from a navigation radar in one embodiment of the present invention.

[0018] Figure 2 Schematic diagram of the resolution of two target echo signals separated by one beam width and with phases changing in 15° steps according to an embodiment of the present invention.

[0019] Figure 3 FIG. 4 is a flow chart of a beam sharpening algorithm according to an embodiment of the present invention.

[0020] Figure 4 Schematic diagram of original echo in azimuth dimension in one embodiment of the present invention.

[0021] Figure 5 Schematic diagram of azimuth echo and dynamic threshold in one embodiment of the present invention.

[0022] Figure 6 FIG. 4 is a schematic diagram of echo segmentation in an embodiment of the present invention.

[0023] Figure 7 Schematic diagram of matching search in one embodiment of the present invention.

[0024] Figure 8 FIG. 1 is a schematic diagram of a residual matrix in an embodiment of the present invention.

[0025] Figure 9 FIG. 1 is a schematic diagram of a sharpened echo 1 according to an embodiment of the present invention.

[0026] Figure 10 FIG. 1 is a schematic diagram of an original echo basis and a sharpened echo basis in one embodiment of the present invention.

[0027] Figure 11 FIG. 2 is a schematic diagram of a sharpened echo 2 according to an embodiment of the present invention.

[0028] Figure 12 Schematic diagram of sea surface echo from a navigation radar in one embodiment of the present invention.

[0029] Figure 13 Schematic diagram of a test scenario in one embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following is a detailed description of an embodiment of the present invention in conjunction with the accompanying drawings: This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and specific operation process are given, but the protection scope of the present invention is not limited to the following embodiment.

[0031] A navigation radar beam sharpening method based on peak-valley detection and matching pursuit is proposed. For a navigation radar, assume that there are two ships at the same detection range of the radar, but with a small azimuth difference of Δθ. If the navigation radar can just determine that the two ships are two different targets instead of treating them as one ship, the azimuth resolution of the navigation radar is called Δθ. The azimuth resolution of the navigation radar is approximately equal to the half-power beamwidth of the navigation radar antenna. The relative phase of the two targets will affect the target resolution process (such as Figure 2 The most difficult to distinguish occurs when the two targets are in phase.

[0032] Figure 3 This is a flowchart of the beam sharpening algorithm based on peak-valley detection and matching search. For targets whose azimuth interval exceeds the half-power beamwidth or whose azimuth echoes form peaks and valleys due to large relative phase differences, peak-valley detection is used to distinguish them. For targets that form a single peak in the azimuth echo, matching search is used to determine whether they are the superposition of two single-target echoes with the same phase. The specific steps are as follows:

[0033] (1) Echo segmentation: First, the azimuth echo is segmented to extract the continuous echo area above the detection threshold; the azimuth echo data is traversed according to the index. In this process, the continuous areas above the pre-set detection threshold and whose length is greater than a specific value are extracted, and the starting positions of these continuous areas are stored.

[0034] (2) Peak and valley detection: Then, a dynamic threshold based on the maximum value point is used to perform peak and valley detection on each continuous echo area. For each continuous area, a dynamic threshold is used to perform peak and valley detection. The dynamic threshold is generated by generating a detection platform slightly lower than the amplitude value of the data point near the neighborhood of each data point. The maximum value of the overlapping detection platforms is taken as the dynamic threshold to avoid random noise interference. The area above the dynamic threshold is the peak area.

[0035] (3) Matching search: Secondly, for a continuous echo region with only one peak, a matching search algorithm is used to further determine whether it is composed of the superposition of two single-target echoes. For the case where there is only one peak in the continuous region, a specific matching search algorithm is used to perform beam sharpening processing, and the least squares method is used to determine whether the single-peak continuous region has the possibility of being decomposed into two superimposed single-target echoes.

[0036] (4) Echo generation: Finally, the matching search results and the sharpened beam pattern are used to generate a sharpened echo, which is then fused with the peak and valley detection results to form an azimuth-dimensional beam-sharpened echo. For continuous regions with more than two peaks, the dynamic threshold detection result is used as the sharpened echo. For continuous regions with only one peak, the sharpened echo is generated using the antenna sharpening pattern based on the two single target positions and amplitudes obtained by the matching search algorithm.

[0037] First, the echo signal of the navigation radar receiver in each PRF is sampled and digitally processed to obtain the distance dimension echo data and store it; after the navigation radar antenna rotates one circle, the azimuth dimension echo data of the same distance is extracted for processing, such as Figure 4 shown.

[0038] Perform threshold detection on the azimuth echo data and extract the continuous area above the threshold. The detection process is as follows: Figure 6 As shown, where Th is the detection threshold, and the storage length is greater than L min The start and end index i of the continuous region a and i b .

[0039] For each continuous region extracted from the original azimuth echo, a dynamic threshold is used for peak and valley detection. The dynamic threshold is generated by traversing each data point and generating a detection platform with an amplitude of A-Δ (where A is the amplitude of the data point and Δ is a fixed deviation, both in dB) in its neighborhood [-L1, L1]. The maximum value of the overlapping detection platforms and the detection threshold Th is taken as the dynamic threshold. The formula is as follows:

[0040]

[0041] Where s[i] is the original echo in the azimuth dimension, and y[i] is the dynamic threshold. Typical dynamic thresholds and their detection results are shown in the following table: Figure 5 As shown in FIG, the use of dynamic threshold for detection can effectively avoid the random noise interference problem in the traditional maximum value detection method.

[0042] After performing peak and valley detection on the echo of each continuous region, the same algorithm as the echo segmentation is used (using a dynamic threshold instead of the fixed detection threshold Th) to extract the sub-continuous region. Each sub-continuous region represents a peak. If there is only one peak, the matching search algorithm is used for beam sharpening. The specific steps are as follows: with the maximum point of the continuous region as the center, two single-target echo bases are generated in the space of [-L, 0] and [0, L] (L is equal to half the length of the continuous region) (the main lobe of the antenna pattern or the actual strong point target echo is used as the reference for translation, denoted as a0(t)); the echo of the continuous region is projected onto the two groups of bases using the least squares method and the residual energy is recorded; the two groups of single-target echo bases are traversed in the space of [-L, 0] and [0, L] respectively, and the two groups of bases with non-negative projection energies and the smallest residual values on the two groups of bases are used as the sharpening results. The algorithm flow is as follows Figure 7 As shown, the residual matrix example is as follows Figure 8 As shown in Figure 2, it can be seen that the minimum value of the residual matrix of the single target echo is located at the origin, while the minimum value of the residual matrix of the two target echoes is located at a non-origin. Note that due to the restriction that the energy of the echo projection on the two single target echo bases is non-negative, a part of the residual matrix is invalid ( Figure 8 medium gray area), Figure 8 a in is a single target echo, Figure 8 The b in the equation is the echo of two targets.

[0043] In actual echoes, in order to prevent the beam sharpening results of adjacent distance units from jumping, the azimuth echoes of adjacent distance units can be taken for joint projection processing. The specific method is to Figure 7 Replace the formula in the middle box ① as follows:

[0044] A=[a1 a2]

[0045] X=(A T A) -1 A T Y

[0046] Y=[y k-M …y k …y k+M ]

[0047] Replace the formula in box ② as follows:

[0048]

[0049] Where k is the index of the current distance unit, M is the distance neighborhood width, X=[x k-M …x k …x k+M ], x k That is the azimuth echo projection result of the current range unit.

[0050] For continuous echoes with two or more peaks, the dynamic threshold peak-valley detection result is used as the sharpened echo, see Figure 9 For continuous echoes with only one peak, the sharpened echo basis a0′(t) (generated by downsampling the original echo basis a(t), a0′(t)=a0(nt), where n is the downsampling multiple, see Figure 10 ) and the least squares projection result of the previous step generate a sharp echo, see Figure 11 , the calculation method is as follows:

[0051] y′(i)=x(1)·a0′(i-i1)+x(2)·a0′(i-i2)

[0052] Where i1 and i2 are the offset values of the optimal projection basis relative to the reference basis, and x(1) and x(2) are the projection amplitudes of the echo energy on the optimal projection basis.

[0053] The test scene and sharpened echo are as follows: Figure 12 and Figure 13 As shown, Figure 12 The a in is the original echo, Figure 12 The sharpening result is shown in Figure b. Targets 1 and 2 are corner reflectors protruding from the water (fixed in place by weights), and target 3 is a small wooden boat carrying a corner reflector. As can be seen, when target 3 approaches target 1, the original radar video echo cannot directly distinguish the two targets, while the sharpening algorithm can effectively distinguish them.

[0054] This method is generally suitable for simple scenes with few targets (such as water surfaces). Pre-processing clutter suppression can reduce the number of targets to be resolved. By utilizing region segmentation and dynamic threshold detection techniques, it can also handle more complex scenes. This method is also applicable to super-resolution processing of radar echoes in other dimensions, such as range and Doppler, such as automotive millimeter-wave radars.

[0055] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A navigation radar beam sharpening method based on peak-valley detection and matching pursuit, characterized in that: For navigation radar, suppose there are two ships at the same detection range of the radar, but their azimuth angles differ by a very small Δθ. If the navigation radar can just determine that the two ships are two different targets, rather than treating them as one ship, then the azimuth resolution of the navigation radar is called Δθ. For targets whose azimuth interval exceeds the half-power beamwidth or whose azimuth echoes form peaks and valleys due to large relative phase differences, the peak-valley detection method is used to distinguish them. For targets that form a single peak in the azimuth echo, a matching search is performed to determine whether it is the superposition of two single target echoes with the same phase. The specific operation steps are as follows: (1) Echo segmentation: First, perform echo segmentation on the azimuth echo to extract the continuous echo area above the detection threshold; (2) Peak and valley detection: Then, a dynamic threshold generated based on the maximum value point is used to perform peak and valley detection on each continuous echo area; (3) Matching search: Secondly, for the continuous echo area with only one peak, a matching search algorithm is used to further determine whether it is composed of the superposition of two single target echoes; (4) Echo generation: Finally, the matching search results and the sharpened beam pattern are used to generate sharpened echoes, which are then fused with the peak and valley detection results to form azimuth-dimensional beam-sharpened echoes.

2. The navigation radar beam sharpening method based on peak-valley detection and matching pursuit according to claim 1, characterized in that: In the step (1), the azimuth echo data are traversed according to the index, and in this process, continuous regions that are higher than a preset detection threshold and whose length is greater than a specific value are extracted, and the starting positions of these continuous regions are stored.

3. The navigation radar beam sharpening method based on peak-valley detection and matching pursuit according to claim 1, characterized in that: In step (2), a dynamic threshold is used to perform peak and valley detection on each continuous area. The dynamic threshold is generated by generating a detection platform slightly lower than the amplitude value of the data point near the neighborhood of each data point. The maximum value of the overlapping detection platforms is taken as the dynamic threshold to avoid random noise interference. The area above the dynamic threshold is the peak area.

4. The navigation radar beam sharpening method based on peak-valley detection and matching pursuit according to claim 1, characterized in that: In step (3), for the case where there is only one peak in the continuous area, a specific matching search algorithm is used to perform beam sharpening processing, and the least squares method is used to determine whether the single-peak continuous area has the possibility of being decomposed into two superimposed single target echoes.

5. The navigation radar beam sharpening method based on peak-valley detection and matching pursuit according to claim 1, characterized in that: In the step (4), if there are more than two peaks in the continuous area, the dynamic threshold detection result is used as the sharpened echo; if there is only one peak in the continuous area, the sharpened echo is generated using the antenna sharpening pattern according to the two single target positions and amplitudes obtained by the matching search algorithm.

6. The navigation radar beam sharpening method based on peak-valley detection and matching pursuit according to claim 1, characterized in that: The azimuth resolution of the navigation radar is approximately equal to the half-power beam width of the navigation radar antenna. The relative phase of two targets will affect the target resolution process, and it is most difficult to resolve when the two targets have the same phase.

Images