Bird flock target identification method based on distance-Doppler comprehensive characteristics

Abstract

The invention discloses a bird flock target identification method based on distance-Doppler comprehensive features, and relates to the technical field of radar target identification, and the method comprises the following steps: S1, detecting a group target, and judging the type of the target according to a detection point; s2, performing preliminary judgment on the target according to the distance-Doppler dispersion characteristics of the detection points; s3, carrying out fluctuation judgment between detection points on the target which is judged according to the distance-Doppler dispersion characteristics; and S4, performing detection point power ratio judgment on the target passing the fluctuation judgment between the detection points. According to the bird flock target identification method based on the distance-Doppler comprehensive features, through distance-Doppler dimension detection point distribution, power fluctuation between detection points, a detection point power ratio and other features are judged, target signals are judged step by step in a cascade mode, and then identification results are output. Therefore, the bird flock target recognition accuracy is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of radar target recognition, and in particular to a method for identifying bird flock targets based on range-Doppler comprehensive features. Background Art

[0002] In the civil aviation airport area, bird strikes are major hidden dangers threatening aviation safety. Compared with ordinary single birds, due to their large number, when a bird flock encounters a passenger plane during flight, the probability of multiple birds hitting key parts of the fuselage and engine simultaneously increases significantly, seriously endangering the lives of passengers on board and the flight safety of the aircraft. However, the flight postures of bird flocks are highly maneuverable, and the echo signals of the densely packed individuals inside are intertwined, making it difficult for radar systems to accurately identify bird flock targets, resulting in problems such as high false alarms and missed alarms. Therefore, improving the radar's ability to identify bird flocks and building a solid defense line for aviation safety have become urgent problems to be solved in the field of radar signal processing.

[0003] The range-Doppler scatter feature, as a means of target characterization, provides a new idea for bird flock identification. During the radar detection of a bird flock, due to the differences in flight speeds and distances among individuals and their relative motion relationship with the radar, a unique scatter pattern will be formed in the range-Doppler domain, while ordinary single bird targets usually show a concentrated distribution feature. At the same time, the relative positions of different individuals in the bird flock with respect to the radar are constantly changing, and there are also obvious differences in the occlusion and scattering characteristics among individuals compared with single bird targets, which can also be used as the main basis for judging bird flock targets.

[0004] Currently, bird flock identification mainly relies on traditional single-dimensional feature analysis methods, including threshold detection based on echo amplitude, motion speed estimation based on Doppler frequency, and clustering analysis based on statistical models. The threshold detection method screens suspected targets by setting an amplitude threshold, but it is easily affected by clutter and cannot distinguish target types; the Doppler analysis method uses frequency shift information to judge the motion direction and speed, but it does not make full use of the internal speed distribution characteristics of the bird flock; the clustering analysis method classifies by statistically analyzing feature similarities, but its performance highly depends on feature selection and prior knowledge. In addition, although deep learning-based methods have shown certain advantages in complex scenarios, they have problems such as insufficient training data and high computational resource requirements.

[0005] Therefore, there is an urgent need for a bird flock identification method that combines range-Doppler comprehensive features to improve the identification accuracy and robustness in complex scenarios through multi-dimensional feature fusion. Summary of the Invention

[0006] The objective of the present invention is to provide a method for identifying bird flock targets based on integrated range-Doppler features. By means of features such as the scattering of detection points in the range-Doppler dimension, the power fluctuation between detection points, and the power ratio of detection points, the target signal is judged step by step in a cascaded manner, and then the recognition result is output, thereby improving the accuracy of bird flock target recognition.

[0007] To achieve the above objective, the present invention provides a method for identifying bird flock targets based on integrated range-Doppler features, including the following steps: S1. Detection of group targets, judging the target type according to the detection points; S2. Making a preliminary judgment on the target according to the range-Doppler scattering features of the detection points; S3. For the targets judged by the range-Doppler scattering features, judging the fluctuation between detection points. The detection points that meet the requirements are marked as suspected bird flock targets and enter S4, and other detection points are marked as non-bird flock targets, and the recognition process is aborted; S4. For the targets judged by the fluctuation between detection points, judging the power ratio of detection points. The detection points that meet the requirements are marked as suspected bird flock targets and enter the subsequent process for judgment, and other detection points are marked as non-bird flock targets, and the recognition process is aborted.

[0008] Preferably, S1 includes the following steps: S11. After the echo signal undergoes pulse compression and coherent accumulation, a range-Doppler matrix is obtained. CFAR is used to detect targets in the range-Doppler dimension respectively. The threshold calculation of CFAR detection is as follows: ; where, is the CFAR detection threshold, is the scaling factor, is the reference window length, is the sample value of the range-Doppler unit within the reference window; S12. Traverse all detection points and perform subsequent step-by-step judgments on each detection point.

[0009] Preferably, in S12, the group targets show a cluster distribution, and the single targets show a discrete distribution.

[0010] Preferably, S2 includes the following steps: S21. Taking the current traversed point as the center, extracting the monitoring point with the largest amplitude in the Doppler dimension corresponding to the current distance unit as the peak point, expressed as: ; where, is the distance unit of the peak point, is the Doppler unit of the peak point; S22. Cluster the main peak point and surrounding detected points as follows: ; Among them, represents the set of detected points obtained after clustering, is the number of detected points in the set; and represent the distance clustering range, and represent the Doppler dimension clustering range; S23. Calculate the distance dimension and Doppler dimension scatter characteristics of the set of detected points where the main peak point is located as follows: ; ; ; Among them, is the distance dimension scatter vector, is the Doppler dimension scatter vector, is the scatter characteristic vector; S24. Determine whether the distance scatter characteristic satisfies: ; Among them, and represent the distance dimension and Doppler dimension scatter thresholds respectively.

[0011] Preferably, S3 includes the following steps: S31. Extract all detected points in the Doppler dimension corresponding to the distance cell where the main peak is located and arrange them according to the size of the Doppler cell: ; Among them, is the number of detected points in the Doppler dimension corresponding to the distance cell where the main peak is located. If , it is marked as a non-bird flock target, otherwise continue to judge; S32. Calculate the average power of adjacent non-main peak detected points, obtain the average power of all signal points between adjacent non-main peak detected points in the Doppler dimension of the currently traversed point, and compare the difference between the two with the fluctuation threshold: ; ; Among them, is the fluctuation characteristic vector, represents the signal power corresponding to the distance-Doppler cell, is the number of Doppler cells between adjacent non-main peak detected points, is the index of the detected point in the Doppler dimension of the target is the signal point index between the target Doppler dimension detection points, is the fluctuation threshold.

[0012] Preferably, S4 includes the following steps: S41. For the set of detection points extracted in S31, extract the point with the maximum signal power among the other detection points except the main peak point, denoted as ; S42. Calculate the power ratio of the main peak point and the point with the second largest power , and compare it with the power ratio threshold; ; ; Among them, is the power ratio feature vector, is the power ratio threshold. The detection points that meet the threshold requirements are marked as flock targets, and other detection points are marked as non-flock targets.

[0013] Therefore, the present invention adopts the above-mentioned method for identifying flock targets based on distance-Doppler comprehensive features, combines the feature differences between individual targets during the flight of the group target, combines the distance-Doppler two-dimensional scattering characteristics of the target reflected echo, and comprehensively uses the fluctuation characteristics and power ratio characteristics of the signal in the clustering result to judge the target. Make full use of multi-dimensional feature fusion to improve the recognition accuracy.

[0014] The technical solution of the present invention will be further described in detail below with reference to the drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flowchart of a method for identifying flock targets based on distance-Doppler comprehensive features according to the present invention; Figure 2 is the RD diagram of the flock target of the present invention; Figure 3 is the RD diagram of a single bird target of the present invention; Figure 4 is the Doppler dimension signal diagram of the flock target of the present invention; Figure 5 is the Doppler signal diagram of a single bird target of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The technical solution of the present invention will be further described below with reference to the drawings and embodiments.

[0017] Unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings as understood by those of ordinary skill in the art to which the present invention pertains. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. Words such as "comprising" or "including" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. Words such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Upper", "lower", "left", "right", etc. are only used to indicate relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0018] Embodiment Please refer to Figures 1-5 , the present invention provides a method for identifying a flock target based on distance-Doppler comprehensive features, and the specific steps include: S1. Detection of group targets.

[0019] S11. After the echo signal undergoes pulse compression and coherent integration, a distance-Doppler matrix is obtained, and CFAR (constant false alarm rate detection) is used to perform target detection on the distance-Doppler dimensions respectively. The threshold calculation for CFAR detection is as follows: ; Wherein, is the CFAR detection threshold, is the scaling factor, is the reference window length, is the sample value of the distance-Doppler unit within the reference window.

[0020] S12. Traverse all detection points, and perform subsequent step-by-step judgments on each detection point.

[0021] When the group target is flying, it shows an overall moving trend. On the premise that the radar range and velocity resolutions meet the requirements, after CFAR detection, multiple detection points will appear on the RD map. The group target shows a cluster distribution, and the single target shows a discrete distribution. The detection points are the main basis for the subsequent judgment process.

[0022] S2. Based on the distance-Doppler scatter characteristics of the detection points, make a preliminary judgment on the target.

[0023] S21. Taking the current traversed point as the center, extract the detection point with the largest amplitude in the Doppler dimension corresponding to the current distance unit as the main peak point, which is expressed as: ; Wherein, is the main peak point distance unit, is the main peak point Doppler unit; S22. Cluster the main peak point and surrounding detection points as follows: ; Among them, represents the set of detection points obtained after clustering, is the number of detection points in the set; and represent the distance clustering range, and represent the Doppler dimension clustering range.

[0024] S23. Calculate the distance dimension and Doppler dimension scatter characteristics of the detection point set where the main peak point is located as follows: ; ; ; Among them, is the distance dimension scatter vector, is the Doppler dimension scatter vector, is the scatter characteristic vector; S24. Judge whether the distance scatter characteristic meets: ; Among them, and respectively represent the distance dimension and Doppler dimension scatter thresholds.

[0025] The detection points whose scatter characteristics meet the threshold are marked as suspected bird flock targets and enter the judgment of the subsequent process, and other detection points are identified as non-bird flock targets, and the recognition process is aborted.

[0026] In a clutter environment, the echo signals of single bird targets are aliased with static clutter signals in the distance dimension and Doppler dimension, forming multiple detection points; there are differences in the distances and flight speeds of individual targets in a bird flock, and multiple detection points are also easily formed in the distance dimension and Doppler dimension. By clustering the detection points and judging the scatter characteristics, single targets and group targets can be preliminarily distinguished.

[0027] S3. Perform fluctuation judgment between detection points for the targets judged by the distance-Doppler scatter characteristics.

[0028] S31. Extract all the detection points in the Doppler dimension corresponding to the distance unit where the main peak is located and arrange them according to the size of the Doppler unit: ; Among them, is the number of Doppler dimension detection points corresponding to the range cell where the main peak is located. If , it is marked as a non-bird flock target; otherwise, continue the judgment.

[0029] S32. Calculate the difference between the two and compare it with the fluctuation threshold: ; ; Among them, is the fluctuation feature vector, represents the signal power corresponding to the range-Doppler cell, is the number of Doppler cells between adjacent non-main peak detection points, is the index of the target Doppler dimension detection point, is the index of the signal points between the target Doppler dimension detection points, is the fluctuation threshold.

[0030] The detection points that meet the threshold are marked as suspected bird flock targets and enter the subsequent process for judgment. Other detection points are marked as non-bird flock targets, and the recognition process is aborted.

[0031] In a static clutter environment, there are usually large differences in power between multiple detection points in the Doppler dimension of a single bird target; the individual velocity distribution of a bird flock target is random, and there is no distribution pattern between adjacent detection points in the Doppler dimension. After coherent integration of multiple targets, the spectrum superposition results in the power between two adjacent detection points being stronger than the noise floor level.

[0032] S4. For the targets judged by the fluctuation between detection points, perform a judgment on the power ratio of the detection points.

[0033] S41. For the set of detection points extracted in step S31, extract the point with the maximum signal power among the other detection points except the non-main peak points, and denote it as .

[0034] S42. Calculate the power ratio of the main peak point and the point with the second-largest power , and compare it with the power ratio threshold ; ; Among them, is the power ratio feature vector, is the power ratio threshold. The detection points that meet the threshold requirements are marked as bird flock targets, and other detection points are marked as non-bird flock targets. The bird flock recognition process ends.

[0035] Most of the individual targets in the bird flock target are of similar volume size, and the difference in their echo intensities is small; for the detection points generated by single-target birds, there is a large gap between the signal intensity and the clutter signal intensity. Using this change can better distinguish between bird flock targets and single-target birds.

[0036] In this example, the X-band phased array radar identifies low-altitude bird flock targets. The radar parameters are as follows.

[0037] Table 1 X-band phased array radar parameters ;

[0038] For different targets, fixed-point collection of bird flock target data and single-bird target data, etc., to obtain multiple groups of measured data, and separately process single-bird targets and bird flock targets to obtain Figures 2-5 , perform steps S1-S4 of this patent on the bird flock target data, and statistically obtain that the correct recognition rate of the bird flock is higher than 80%, while the single-bird target data is not recognized as a bird flock. Thus, it proves the accuracy and effectiveness of this method.

[0039] Therefore, the present invention adopts the above-mentioned method for identifying bird flock targets based on distance-Doppler comprehensive features, combines the characteristic differences between individual targets during the flight of the group target, combines the distance-Doppler two-dimensional scattering characteristics of the target reflected echo, and comprehensively utilizes the fluctuation characteristics and power ratio characteristics of the signals in the clustering results to judge the target, making full use of multi-dimensional feature fusion, thereby improving the recognition accuracy.

[0040] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify or equivalently replace the technical solutions of the present invention, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for identifying bird flock targets based on the comprehensive features of distance-Doppler, characterized in that, It includes the following steps: S1. Group target detection, and judge the target type according to the detection points; S2. Make a preliminary judgment on the target according to the range-Doppler spread characteristics of the detection points; S3. For the targets judged by the range-Doppler spread characteristics, judge the fluctuations between detection points. The detection points that meet the requirements are marked as suspected bird flock targets and enter S4, and other detection points are marked as non-bird flock targets, and the recognition process is aborted; S4. For the targets judged by the fluctuations between detection points, judge the power ratio of the detection points. The detection points that meet the requirements are marked as suspected bird flock targets and enter the subsequent process judgment, and other detection points are marked as non-bird flock targets, and the recognition process is aborted.

2. The method for identifying a flock target based on the comprehensive distance-Doppler characteristics according to claim 1, wherein S1 includes the following steps: S11. The echo signal is subjected to pulse compression and coherent accumulation to obtain a range-Doppler matrix, and CFAR is used to detect targets in the range-Doppler dimensions respectively. The threshold calculation of CFAR detection is as follows: ； Among them, is the CFAR detection threshold, is the scaling factor, is the reference window length, is the sample value of the range-Doppler cell within the reference window; S12. Traverse all detection points and perform subsequent hierarchical judgments on each detection point.

3. The method for identifying bird flock targets based on the comprehensive distance-Doppler characteristics according to claim 2, wherein: In S12, group targets show a cluster distribution, and single targets show a discrete distribution.

4. The method for identifying a flock target based on the comprehensive distance-Doppler characteristics according to claim 3, wherein S2 includes the following steps: S21. Take the current traversed point as the center, and extract the monitoring point with the largest amplitude in the Doppler dimension corresponding to the current range cell as the main peak point, which is expressed as: ； Among them, is the main peak point distance unit, is the main peak point Doppler unit; S22. Cluster the main peak point and the surrounding detection points, and the method is as follows: ； Among them, represents the set of detection points obtained after clustering, is the number of detection points in the set; and represents the distance clustering range, and represents the Doppler dimension clustering range; S23. Calculate the spread characteristics of the range dimension and the Doppler dimension for the detection point set where the main peak point is located as follows: ； ； ； Among them, is the range dimension scatter vector, is the Doppler dimension scatter vector, is the scatter feature vector; S24. Determine the distance distribution characteristics Whether it satisfies: ； Among them, and respectively represent the range and Doppler spread thresholds.

5. The method for identifying a flock target based on the comprehensive distance-Doppler characteristics according to claim 4, characterized in that, S3 includes the following steps: S31. Extract all detection points in the Doppler dimension corresponding to the range cell where the main peak is located, and arrange them according to the size of the Doppler cell: ； Among them, is the number of detection points in the Doppler dimension corresponding to the range cell where the main peak is located. If , it is marked as a non-bird flock target; otherwise, continue to make a judgment. S32. Calculate the average power of adjacent non-main peak detection points, and obtain the average power of all signal points between adjacent non-main peak detection points in the Doppler dimension signal of the current traversed point, and calculate the difference between the two and compare it with the fluctuation threshold: ； ； Among them, is the fluctuation feature vector, represents the signal power corresponding to the range-Doppler cell, is the number of Doppler cells between adjacent non-main peak detection points, is the target Doppler dimension detection point index, is the signal point index between target Doppler dimension detection points, is the fluctuation threshold.

6. The method for identifying bird flock targets based on the comprehensive range-Doppler characteristics according to claim 5, characterized in that S4 includes the following steps: S41. For the set of detection points extracted in S31, extract the point with the maximum signal power among the other detection points except the non-main peak points, denoted as ; S42. Calculate the main peak point and the point with the second-largest power value to calculate the power ratio, and compare it with the power ratio threshold; ； ； Among them, is the power ratio feature vector, is the power ratio threshold. Detection points that meet the threshold requirements are marked as flock targets, and other detection points are marked as non-flock targets.

Images