A Detection Method for Dim and Small Targets in a Wide-Area Sparse Array

Abstract

The present invention discloses a method for detecting weak targets in a wide-area sparse array, which relates to the field of radar target detection. The method first constructs a wide-area sparse array and sets a threshold; then, beamforming is performed under multiple targets in the wide-area sparse array; secondly, the largest target in the pattern is found; then, the pattern function of multiple targets is subtracted from the pattern function of the largest target; finally, it is judged according to the threshold and decided whether to repeat the previous two steps until all targets are found. The present invention can solve the problem that small targets are submerged by high sidelobes of large targets when the wide-area sparse array generates high sidelobes due to too large element spacing.

Description

Technical Field

[0001] The present invention relates to the fields of beamforming and radar target detection, and specifically to a method for detecting weak targets in a wide-area sparse array, which can solve the problem that large target high sidelobes submerge small targets when there are high sidelobes due to too large element spacing in a wide-area sparse array. Background Technique

[0002] The high sidelobes caused by too large element spacing in a wide-area sparse array seriously affect the target detection effect. Especially in the case of multiple targets, the high sidelobes of large targets will submerge the main lobes of small targets, resulting in difficulty in accurately detecting small targets. By adopting various array layout optimization algorithms and beamforming algorithms, the sidelobes of the pattern can be reduced, but there are also limitations.

[0003] Most of the current array layout optimization algorithms and beamforming algorithms are applied to uniform arrays or arrays with element spacing slightly larger than about half a wavelength. However, for a wide-area sparse array, the element spacing is too large, and the effect of reducing sidelobes by using the above methods is very limited, and there are still high sidelobes, resulting in difficulty in detecting small targets.

[0004] It can be seen that if only the array layout optimization algorithm or beamforming algorithm for reducing sidelobes is adopted, it is difficult to distinguish between large and small targets in the detection process of a wide-area sparse array. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for detecting weak targets in a wide-area sparse array, which avoids the deficiencies in the above background technique. This method combines the pattern functions at different frequency points in a coherent form, so that the gain of the sidelobes is greatly reduced while the main lobe gain remains unchanged, which can reduce the adverse effects of high sidelobes of the sparse array on detection and is beneficial to the detection of small targets.

[0006] The purpose of the present invention is achieved as follows:

[0007] A method for detecting weak targets in a wide-area sparse array includes the following steps:

[0008] Step 1, construct a wide-area sparse array, determine the coordinates (x m , y m , z m ) of the mth element, and set a threshold TH;

[0009] Step 2, perform beamforming on the wide-area sparse array to obtain the pattern of multiple targets;

[0010]

[0011] ΔR = x m cosθcosφ + y m cosθsinφ + zm sinθ

[0012] ΔR0 = x m cosθ0cosφ0 + y m cosθ0sinφ0 + z m sinθ0

[0013] where F N (φ,θ) is the pattern function with the direction (θ0,φ0), A is the amplitude weighting coefficient, c is the speed of light, f is the operating frequency, φ is the azimuth angle, θ is the elevation angle, and M is the number of array elements;

[0014] Step 3: Find the target with the largest amplitude in the pattern and estimate the position (θ i , φ i ) and the amplitude A i , and calculate the pattern function of this target according to the following formula;

[0015]

[0016] ΔR = x m cosθcosφ + y m cosθsinφ + z m sinθ

[0017] ΔR i = x m cosθ i cosφ i + y m cosθ i sinφ i + z m sinθ i

[0018] where F i (φ,θ) is the pattern function of the largest target found in the current pattern, and i is the serial number of the target found;

[0019] Step 4: Subtract the pattern function of the target found in Step 3 from the pattern function of the multiple targets in Step 2 to obtain a new pattern function F N-1 (φ,θ);

[0020] F N-1 (φ,θ) = F N (φ,θ) - F i (φ,θ)

[0021] Step 5: Obtain the radiation pattern according to the new radiation pattern function and make a judgment based on the threshold. If the maximum value in the radiation pattern is greater than the threshold TH, repeat Steps 3-4; if the maximum value in the radiation pattern is less than the threshold TH, exit the loop and complete the detection of weak and small targets.

[0022] Step 1: Construct a wide-area sparse array, determine the coordinates (x m , y m , z m ) of the m-th array element, and set the threshold TH of the spatial CLEAN algorithm.

[0023] Step 2: Perform beamforming on the wide-area sparse array to obtain the radiation patterns of multiple targets.

[0024]

[0025] ΔR = x m cosθcosφ + y m cosθsinφ + z m sinθ

[0026] ΔR0 = x m cosθ0cosφ0 + y m cosθ0sinφ0 + z m sinθ0

[0027] where F N (φ,θ) is the radiation pattern function with the pointing direction of (θ0,φ0), A is the amplitude weighting coefficient, c is the speed of light, f is the operating frequency, φ is the azimuth angle, θ is the elevation angle, and M is the number of array elements.

[0028] Step 3: Find the target with the largest amplitude in the radiation pattern, estimate the position (θ i , φ i ) and the amplitude A i , and calculate the radiation pattern function of this target according to the following formula;

[0029]

[0030] ΔR = x m cosθcosφ + y m cosθsinφ + z m sinθ

[0031] ΔR i = x m cosθ i cosφ i + y m cosθ i sinφ i + z m sinθi

[0032] Among them, F i (φ,θ) is the directional function of the largest target found in the current directional map, and i is the serial number of the target found;

[0033] Step 4: Subtract the directional pattern function of the target found in step 3 from the directional pattern function of the multi-target in step 2 to obtain a new directional pattern function F N-1 (φ,θ);

[0034] F N-1 (φ,θ)=F N (φ,θ)-F i (φ,θ)

[0035] Step 5: In the new directional map, make a judgment based on the threshold: if the maximum value in the directional map is greater than TH, repeat steps 3 to 4; if the maximum value in the directional map is less than TH, exit the loop and complete the sparse array spatial domain CLEAN algorithm;

[0036] Compared with the background technology, the present invention has the following advantages:

[0037] 1. The present invention finds the large and small targets in the directional pattern of multiple targets in the airspace in an iterative manner, and subtracts their main lobes and side lobes. The directional pattern after subtracting the side lobes of the large target has a lower side lobe level, which is more beneficial to the detection of small targets.

[0038] 2. The present invention has certain significance for target detection of wide-area sparse arrays. Especially in the case of multi-target detection, the array optimization algorithm and the beamforming algorithm have very limited effect on reducing the side lobes of the wide-area sparse array pattern. The method of the present invention eliminates the side lobes of large targets in an iterative manner, thereby more effectively reducing the adverse effects of the high side lobes of large targets on weak target detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a flow chart of a method for detecting small targets with a wide-area sparse array in an embodiment of the present invention.

[0040] Figure 2 It is the directional pattern of a small target appearing in the side lobe of a large target.

[0041] Figure 3 is the estimated direction map of the large target.

[0042] Figure 4 It is the directional pattern function minus the estimated large target directional pattern function.

[0043] Figure 5 is the estimated direction map of the small target.

[0044] Figure 6 It is the radiation pattern after subtracting small targets. Specific embodiments

[0045] A method for detecting weak and small targets in a wide-area sparse array. First, a wide-area sparse array is constructed and a threshold is set. Then, beamforming is performed for multiple targets in the wide-area sparse array. Next, the largest target in the radiation pattern is found. Then, the radiation pattern function of the multiple targets is subtracted from the radiation pattern function of the largest target. Finally, it is judged according to the threshold whether to repeat the previous two steps until all targets are found. As Figure 1 shown, it specifically includes the following steps:

[0046] Step 1: Construct a wide-area sparse array, determine the coordinates (x m , y m , z m ) of the m-th array element, and set the threshold TH of the CLEAN algorithm in the spatial domain;

[0047] Among them, the threshold TH is used to ensure the normal operation of the CLEAN algorithm;

[0048] Step 2: Perform beamforming on the wide-area sparse array to obtain the radiation pattern of multiple targets;

[0049]

[0050] ΔR = x m cosθcosφ + y m cosθsinφ + z m sinθ

[0051] ΔR0 = x m cosθ0cosφ0 + y m cosθ0sinφ0 + z m sinθ0

[0052] Among them, F N (φ, θ) is the radiation pattern function with the pointing direction of (θ0, φ0), A is the amplitude weighting coefficient, c is the speed of light, f is the operating frequency, φ is the azimuth angle, θ is the elevation angle, and M is the number of array elements;

[0053] Among them, the radiation pattern function is a complex function, and the radiation pattern can be obtained only by taking the modulus, that is |·| represents taking the modulus;

[0054] Step 3: Find the target with the largest amplitude in the radiation pattern, estimate the position (θ i , φ i ) and the amplitude A i of this point, and calculate the radiation pattern function of this target according to the following formula;

[0055]

[0056] ΔR = x m cosθcosφ + y m cosθsinφ + z m sinθ

[0057] ΔR i = x m cosθ i cosφ i + y m cosθ i sinφ i + z m sinθ i

[0058] where F i (φ,θ) is the pattern function of the maximum target found in the current pattern, and i is the serial number of the found target;

[0059] Step 4, Subtract the pattern function of the target found in Step 3 from the pattern function of the multi-target in Step 2 to obtain a new pattern function F N-1 (φ,θ);

[0060] F N-1 (φ,θ) = F N (φ,θ) - F i (φ,θ)

[0061] Step 5, In the new pattern, make a judgment according to the threshold: If the maximum value in the pattern is greater than TH, repeat Steps 3 - 4; If the maximum value in the pattern is less than TH, exit the loop and complete the CLEAN algorithm for the sparse array airspace;

[0062] The effect of this method can be further illustrated by the following simulation experiments:

[0063] 1. Experimental scenario:

[0064] Now, a wide-area sparse linear array is constructed. The number of array elements M = 20, the array aperture is 1900 m, and the positions of the array elements are (0,0,0), (353.6 m,0,0), (414.1 m,0,0), (576.1 m,0,0), (751.4 m,0,0), (762.3 m,0,0), (773.9 m,0,0), (999.2 m,0,0), (1000.1 m,0,0), (1038.6 m,0,0), (1041.4 m,0,0), (1277.5 m,0,0), (1287.9 m,0,0), (1330.6 m,0,0), (1419.4 m,0,0), (1483.0 m,0,0), (1660.0 m,0,0), (1743.3 m,0,0), (1778.3 m,0,0), (1900 m,0,0), and the operating frequency is 2 GHz.

[0065] 2. Simulation content:

[0066] Set that there is a large target and a small target at 0° and 0.02353° respectively in the spatial domain. The small target is located at the sidelobe of the large target. For this, the spatial CLEAN algorithm is used to process the large and small targets successively. Among them, the signal-to-noise ratio SNR of the large target is 20, the signal-to-noise ratio SNR of the small target is 12, and the threshold of the CLEAN algorithm is set to -35 dB.

[0067] 3. Analysis of simulation results:

[0068] From Figure 2 it can be seen that the small target appears at the sidelobe of the large target. Due to the high sidelobe of the large target, it is difficult to distinguish other targets except the large target in the direction diagram, that is, the small target is submerged by the sidelobe of the large target;

[0069] From Figure 3 it can be seen that the position estimation of the large target is -0.00003°, and the estimation error is very small;

[0070] From Figure 4 it can be seen that in the direction diagram after subtracting the estimated large target, the interference of the high sidelobe of the large target is removed, and the small target becomes very clear. And at this time, the maximum amplitude in the direction diagram is about -20 dB, which is greater than the set threshold of -35 dB, so the iteration continues;

[0071] From Figure 5 it can be seen that the position estimation of the small target is 0.0236°, and the estimation error is 1%;

[0072] From Figure 6It can be seen that after subtracting the pattern function of the estimated small target from the pattern function, the maximum amplitude in the pattern is approximately -43 dB at this time, which is less than the set threshold of -35 dB. Therefore, the iteration of the CLEAN algorithm is terminated at this time, and finally two targets are found.

Claims

1. A method for detecting weak and small targets in a wide-area sparse array, characterized in that Including the following steps: Step 1, construct a wide-area sparse array, determine the coordinates (x m , y m , z m ) of the m-th array element, and set the threshold TH; Step 2, wide-area sparse array beamforming to obtain the direction diagrams of multiple targets; ΔR = x m cosθcosφ + y m cosθsinφ + z m sinθ ΔR0 = x m cosθ0cosφ0 + y m cosθ0sinφ0 + z m sinθ0 Among them, F N (φ,θ) is the pattern function with the direction (θ0,φ0), A is the amplitude weighting coefficient, c is the speed of light, f is the operating frequency, φ is the azimuth angle, θ is the elevation angle, and M is the number of array elements; Step 3, find the target with the largest amplitude in the radiation pattern and estimate the position of this point ($\theta$ i , $\varphi$ i ) and amplitude $A$ i , and calculate the radiation pattern function of this target according to the following formula; ΔR = x m cosθcosφ + y m cosθsinφ + z m sinθ ΔR i = x m cosθ i cosφ i + y m cosθ i sinφ i + z m sinθ i Among them, F i (φ,θ) is the pattern function of the maximum target found in the current pattern, and i is the serial number of the found target; Step 4: Subtract the direction pattern function of the target found in Step 3 from the multi-target direction pattern function in Step 2 to obtain a new direction pattern function F N-1 (φ,θ); F N-1 (φ,θ) = F N (φ,θ) - F i (φ,θ) Step 5, obtain the direction diagram according to the new direction diagram function and make a judgment based on the threshold: if the maximum value in the direction diagram is greater than the threshold TH, repeat steps 3 to 4; if the maximum value in the direction diagram is less than the threshold TH, exit the loop and complete the detection of weak and small targets.

Images