A non-uniform sea clutter target detection method based on detection performance evaluation

Abstract

The present invention relates to the field of radar sea target detection, and discloses a non-uniform sea clutter target detection method based on detection performance evaluation. By constructing a multi-level CFAR detection model based on sea clutter characteristics, azimuth automatic partitioning, range-segmented target detection, and real-time evaluation of the detection performance of partitioned targets are carried out. Combining the evaluation results, a partitioned automatic threshold control method under multiple constraints is adopted to adjust the target detection threshold levels of different partitions. The present invention constructs a multi-level CFAR detection model based on sea clutter characteristics, introduces a feedback mechanism of detection performance evaluation into the adjustment of detection parameters, realizes automatic threshold control of the detector, and improves the target detection performance in a non-uniform sea clutter environment; moreover, the detection performance evaluation is convenient to implement and has a low algorithm complexity. Through detection performance feedback, it avoids the deterioration of detection performance caused by non-uniform and non-Gaussian sea clutter environments under complex sea conditions, improves the target detection performance under complex sea conditions, and has strong robustness.

Description

Technical Field

[0001] The present invention relates to the field of radar sea target detection, and particularly to a non-uniform sea clutter target detection method based on detection performance evaluation. Background Art

[0002] The constant false alarm rate (CFAR) detection technology is a technology in which a radar system discriminates the signal and noise output by a receiver under the condition of keeping the false alarm probability constant to determine whether a target signal exists.

[0003] Sea clutter is typical dynamic surface clutter. Its complex motion characteristics and composition result in complex sea clutter characteristics. It is colored non-Gaussian clutter, and the theoretical detection threshold under Gaussian background is extremely prone to false alarms. Moreover, sea clutter is greatly affected by sea conditions. The echo power and statistical characteristics of sea clutter under different sea conditions vary greatly, and fixed detectors are difficult to adapt to different sea conditions. In addition, the echo power of sea clutter varies greatly under different wind directions, showing obvious spatial non-uniformity. When the same detector detects in the crosswind area and the headwind area, the number of output targets and false alarms has obvious differences. To address the above problems, generally, by segmenting the distance and partitioning the azimuth, a locally uniform environment is constructed, and the sea clutter characteristics are sensed in real time to achieve adaptive adjustment of detection parameters. This method has the following disadvantages:

[0004] 1. Large data calculation amount and complex algorithm. Due to the colored non-Gaussian nature and spatial non-uniformity of sea clutter, real-time sensing of sea clutter characteristics requires estimating the statistical characteristics of sea clutter by partitioning and segmenting the distance, resulting in a large data calculation amount, which brings difficulties to system implementation; especially when searching over a large range at long distances, the real-time performance of feature estimation cannot be guaranteed.

[0005] 2. Poor accuracy of clutter feature estimation results and susceptibility to interference. The effect of real-time sensing of sea clutter feature estimation is affected by the number of samples. The more samples, the more accurate the estimation result. However, due to partitioning and segmenting the distance, the number of estimated samples decreases, and it is extremely susceptible to the influence of ground clutter, targets, and interference echoes, resulting in serious deviations in the real-time estimation of sea clutter characteristics. The detection parameters adjusted based on this may even lead to deterioration of detection performance. Summary of the Invention

[0006] In view of this, the present invention provides a non-uniform sea clutter target detection method based on detection performance evaluation, which can achieve stable automatic threshold control through feedback adjustment of a small scanning range without parameter estimation, so it is not easily affected by ground clutter, targets, and interference, and has strong robustness.

[0007] A non-uniform sea clutter target detection method based on detection performance evaluation includes the following steps:

[0008] Step 1: Constrained by the radar system parameters, construct a CFAR detection model based on the characteristics of sea clutter in multi-level sea states, and partition the radar scanning angle range;

[0009] Step 2: Perform CFAR detection based on distance-segment thresholds on the echo of the current radar pulse in each angle partition, and record the number of CFAR threshold-crossing points;

[0010] Step 3: Statistically analyze the saturation rate and false alarm rate of the number of CFAR threshold-crossing points of the pulses in each angle partition;

[0011] Step 4: Adjust the automatic threshold level of each angle partition according to the saturation rate and false alarm rate of each angle partition.

[0012] Furthermore, the constraint parameters of the radar system in Step 1 include range, radar beam width, resolution, and radar transmit power.

[0013] Furthermore, the statistical process of the saturation rate SatR in Step 3 is as follows:

[0014]

[0015] where M is the upper limit of the number of targets output by the current pulse, N is the number of distance segments divided by the current pulse at a certain distance interval, then the saturation number of each distance segment is M / N, and i n is the number of CFAR threshold-crossing points in the nth distance segment.

[0016] Furthermore, the statistical method of the false alarm rate Rf in Step 3 is as follows:

[0017]

[0018] where R is the radar range, Θ is the degree of the corresponding angle partition, θ is the radar beam width, and ΔR is the radar range resolution.

[0019] Furthermore, in Step 4, according to the geographical system angle partition, saturation rate, false alarm rate of the current scanning range, as well as the preset upper and lower threshold values of the saturation rate and false alarm rate, a decision is made, and according to the decision result, the threshold level for the next scan of the corresponding angle partition is adjusted for parameter calling during the next scan.

[0020] Furthermore, the decision basis is:

[0021] When the saturation rate is greater than the upper limit Th Sat_up or the false alarm rate is greater than the upper limit Th Rf_up , the threshold level is increased by one level; when the saturation rate is less than the lower limit Th Sat_down or the false alarm rate is less than the lower limit Th Rf_down , the threshold level is decreased by one level;

[0022] In other cases, the threshold level remains unchanged.

[0023] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0024] 1. The detection performance evaluation of the present invention is easy to implement and has a low algorithm complexity. Compared with the method of real-time sensing of sea clutter, the method of the present invention is based on data-level processing, carried out according to the scanning range, easy to implement and with less data processing volume, and the computational complexity is reduced;

[0025] 2. The present invention constructs a multi-level CFAR detection model based on sea clutter characteristics, and adaptively adjusts the multi-level threshold with the detection result performance as feedback information, realizing automatic threshold control. Since parameter estimation is not required, it is not easily affected by ground clutter, targets and interference. Through detection performance feedback, the system has strong robustness and is easy to converge; moreover, stable automatic threshold control can be achieved through feedback adjustment of a small number of scanning ranges. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0027] Figure 1 It is a flowchart of the non-uniform sea clutter target detection method based on detection performance evaluation in Embodiment 2;

[0028] Figure 2 It is a schematic diagram of the detection threshold level setting in Embodiment 2;

[0029] Figure 3 It is a schematic diagram of the correspondence between geographical system angles and pulse numbers in Embodiment 2;

[0030] Figure 4 It is a graph of the false alarm rate of all pulses in the current row in Embodiment 2;

[0031] Figure 5 It is a schematic diagram of the feedback coefficient of the current scanning range in Embodiment 2;

[0032] Figure 6 It is a schematic diagram of the partition of the scanning angle range of the current scanning row in Embodiment 2. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0033] The embodiments of the present application will be described in detail below with reference to the drawings.

[0034] The following describes the implementation modes of the present application through specific examples. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The present application can also be implemented or applied through other different specific implementation modes. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, without conflict, the following embodiments and the features in the embodiments can be combined with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts belong to the scope of protection of the present application.

[0035] Embodiment 1

[0036] A non-uniform sea clutter target detection method based on detection performance evaluation includes the following steps:

[0037] Step 1: Constrained by radar system parameters, construct a CFAR detection model based on multi-level sea clutter characteristics, and partition the radar scanning angle range.

[0038] Step 2: Perform CFAR detection based on distance-segmented thresholds on the echo of the current radar pulse in each angle partition, and record the number of CFAR threshold-crossing points.

[0039] Step 3: Statistically analyze the saturation rate and false alarm rate of the number of CFAR detection threshold-crossing points of the pulses in each angle partition.

[0040] Step 4: Adjust the automatic threshold level of each angle partition according to the saturation rate and false alarm rate of each angle partition.

[0041] In this embodiment, first, constrained by radar system parameters, construct a multi-level CFAR detection model based on sea clutter characteristics; then, for spatial non-uniformity, perform distance-segmented target detection through azimuth automatic partitioning; secondly, adopt a target detection performance evaluation method based on processing capacity constraints to evaluate the target detection performance in real time by partition; finally, combined with the evaluation results of the partition target detection performance, adopt a partition automatic threshold control method under multiple constraints to adjust the target detection threshold levels of different partitions. This embodiment realizes automatic threshold control of the detector and improves the target detection performance in a non-uniform sea clutter environment by constructing a multi-level CFAR detection model based on sea clutter characteristics and introducing a feedback mechanism of detection performance evaluation into detection parameter adjustment; moreover, the detection performance evaluation is easy to implement, the algorithm complexity is low, and through detection performance feedback, it avoids the deterioration of detection performance caused by non-uniform and non-Gaussian sea clutter environments under complex sea conditions, improves the target detection performance under complex sea conditions, and has strong robustness.

[0042] Example 2

[0043] See Figure 1-6 , in this embodiment, the radar range is 40 km, the radar beam width is 1.5°, and the range resolution is 20 m.

[0044] Step 1, establish a multi-level CFAR detection parameter model

[0045] The echo power and statistical characteristics of sea clutter under different sea conditions vary greatly, and a fixed detector is difficult to adapt to different sea conditions; in addition, the background environmental clutter-to-noise ratios at different distances are different. The clutter-to-noise ratio is relatively high at close range, and the background environment is jointly affected by clutter and noise, mainly by clutter. The clutter-to-noise ratio is relatively low at long range, and the background environment is mainly affected by noise. Therefore, constrained by the radar system parameters, based on the radar equation and the signal-to-clutter-plus-noise ratio relationship under different sea state levels, a multi-level range-segmented CFAR detection parameter model based on sea state level mapping is established, and the detection parameter levels are initialized.

[0046] In this embodiment, according to sea state levels 1 to 5, the CFAR detection parameters are set to levels 1 to 5 (in this embodiment, the detection threshold coefficient increases step by step with the threshold level), as Figure 2 shown; initialize the detection parameter level 3;

[0047] Step 2, automatic partitioning of non-uniform environment

[0048] The echo power of sea clutter varies greatly under different wind directions, showing obvious spatial non-uniformity. The scanning angle range is partitioned according to certain rules, and a local clutter-uniform environment is constructed within each partition. The detection angle range can be divided into the upwind area and the crosswind area according to prior meteorological or sea state prediction information; it can also be divided into different angle areas at equal intervals according to the angle uniform partitioning principle.

[0049] Figure 6 is a schematic diagram of the scanning angle range partitioning for the current scanning line; in this embodiment, according to the uniform angle partitioning criterion, the geographical azimuth angle is partitioned at 3° intervals, and a total of 120 areas are divided in the 360° range, as Figure 3 shown; the radar performs azimuth scanning, and each pulse corresponds to a geographical angle, and the processing is carried out pulse by pulse.

[0050] Step 3, range-segmented target detection

[0051] Perform constant false alarm (CFAR) detection based on range-segmented thresholds on the echo of the current radar pulse, and record the number of CFAR threshold crossing points.

[0052] Step 4, detection performance evaluation based on saturation rate and false alarm rate statistics

[0053] Statistically analyze the number of pulses passing through the threshold points in each angular partition of the current scanning range for the saturation rate and false alarm rate, and form a statistical curve graph of the angular partition in the geographical system of the current scanning range, the detection saturation rate, and the false alarm rate, as Figure 4 shown.

[0054] The method for statistically analyzing the saturation rate is as follows: The upper limit of the current pulse output target is M, and N distance segments are divided at a certain distance interval. Then the saturation number of each distance segment is M / N, and the number of CFAR threshold-crossing points in the nth distance segment is i n , then the saturation rate SatR of the current pulse is:

[0055]

[0056] In this embodiment, the method for statistically analyzing the saturation rate is as follows: Taking a certain angular partition as an example, the upper limit of the single-pulse output target is M = 1024, and the current range is divided into N = 16 distance segments. Then the saturation number of each distance segment M / N is 64, and the number of CFAR threshold-crossing points i n in a certain distance segment is 8. Then the saturation rate of this distance segment is 12.5%, and the saturation rate of the current pulse is:

[0057]

[0058] The method for statistically analyzing the false alarm rate is as follows: The false alarm rate statistics is that, according to the range R, the angular partition Θ, the radar beam width θ, and the resolution ΔR, the total number of radar detection units is calculated. Among all the threshold-crossing points M detected by CFAR, the measurement points except for the stable threshold-crossing points N are false alarm points, and the ratio of the number of false alarm points to the total number of radar detection units is the false alarm rate. The false alarm rate Rf of the current pulse is:

[0059]

[0060] Taking a certain angular partition as an example, when the range is 100 km and the angular partition interval is 3°, the radar beam width is 1.5° and the resolution is 20 m, then the total number of radar detection units is (3 / 1.5)*(100000 / 20) = 10000. Among the 100 threshold-crossing points detected by CFAR, the number of stable threshold-crossing points is 60. Then the false alarm rate of the current partition is 4e-3.

[0061] Step 5, Automatic threshold control for partitions under multiple constraints

[0062] Construct an automatic threshold control model based on detection performance constraints, make a decision according to the angular partition in the geographical system of the current scanning range and the saturation rate / false alarm rate in Step 4, and obtain the feedback coefficient of the current scanning range, as Figure 5 shown; then adjust the threshold of the corresponding angular partition in the geographical system during the next scan according to the feedback coefficient for parameter call during the next scan. The adjustment method is the feedback coefficient K fWhen it is equal to 1, the threshold is raised by one level; feedback coefficient K f When it is equal to -1, the threshold is lowered by one level; feedback coefficient K f When it is equal to 0, the threshold remains unchanged.

[0063] The decision is made according to the saturation rate / false alarm rate and the current scan range feedback coefficient K is obtained f is as follows:

[0064]

[0065] In this embodiment, the upper limit Th Sat_up and the lower limit Th Sat_down of the saturation rate are respectively set to 80% and 20%, and the upper limit Th Rf_up and the lower limit Th Rf_down of the false alarm rate are respectively set to 1e-3 and 1e-6.

[0066] That is, when the saturation rate is greater than the upper limit Th Sat_up or the false alarm rate is greater than the upper limit Th Rf_up , the feedback coefficient K f is set to 1; when the saturation rate is less than the lower limit Th Sat_down or the false alarm rate is less than the lower limit Th Rf_down , the feedback coefficient K f is set to -1; in other cases, the feedback coefficient K f is set to 0.

[0067] Then, according to the feedback coefficient, the threshold of the corresponding geographical system angle partition during the next line scan is adjusted for parameter call during the next line scan. Feedback coefficient K f When it is equal to 1, the threshold is raised by one level; feedback coefficient K f When it is equal to -1, the threshold is lowered by one level; feedback coefficient K f When it is equal to 0, the threshold remains unchanged.

[0068] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in this application should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A non-uniform sea clutter target detection method based on detection performance evaluation, characterized in that It includes the following steps: Step 1: Constrained by radar system parameters, construct a CFAR detection model based on multi-level sea clutter characteristics and partition the radar scanning angle range; Step 2: Perform CFAR detection based on distance-segmented thresholds on the echo of the current pulse of the radar in each angle partition, and record the number of CFAR threshold-crossing points; Step 3: Statistically analyze the saturation rate and false alarm rate of the number of CFAR detection threshold-crossing points of the pulses in each angle partition; Step 4: Adjust the automatic threshold level of each angle partition according to the saturation rate and false alarm rate of each angle partition.

2. The non-uniform sea clutter target detection method based on detection performance evaluation according to claim 1, wherein: The constraint parameters of the radar system in Step 1 include range, radar beam width, resolution, and radar transmit power.

3. The non-uniform sea clutter target detection method based on detection performance evaluation according to claim 1, characterized in that: The statistical process of the saturation rate SatR in Step 3 is as follows: Where M is the upper limit number of the current pulse output target, N is the number of distance segments into which the current pulse is divided at a certain distance interval, then the saturation number of each distance segment is M / N, and i n is the number of CFAR threshold-crossing points in the nth distance segment.

4. The non-uniform sea clutter target detection method based on detection performance evaluation according to claim 3, characterized in that: The statistical method of the false alarm rate Rf in Step 3 is as follows: Where R is the distance range of the radar, Θ is the degree of the corresponding angle partition, θ is the radar beam width, and ΔR is the radar range resolution.

5. The non-uniform sea clutter target detection method based on detection performance evaluation according to claim 1, wherein: In Step 4, based on the geographical system angle partition, saturation rate, false alarm rate of the current scanning range, as well as the preset upper and lower threshold values of the saturation rate and false alarm rate, a judgment is made. According to the judgment result, the threshold level for the next line of scanning in the corresponding angle partition is adjusted for parameter call during the next line of scanning.

6. The non-uniform sea clutter target detection method based on detection performance evaluation according to claim 5, characterized in that: The judgment basis is: When the saturation rate is greater than the upper limit Th Sat_up or the false alarm rate is greater than the upper limit Th Rf_up , the threshold level is increased by one level; When the saturation rate is less than the lower limit Th Sat_down or the false alarm rate is less than the lower limit Th Rf_down , the threshold level is lowered by one level; In other cases, the threshold level remains unchanged.

Images