A radar point track condensation method based on echo envelope detection

Abstract

The present invention discloses a radar point condensation method based on echo envelope detection, which comprises the following steps: receiving radar original detection point data; merging the detection points to form a plurality of point ambiguity regions; determining whether each point ambiguity region meets waveform resolution; performing azimuth resolution processing on the point ambiguity regions that meet waveform resolution to form one or more "small cluster" regions after resolution; performing range resolution processing on each "small cluster" region formed by azimuth resolution to form the final respective point regions; calculating the point positions condensed by each region according to the amplitude weighting method for the point regions, and outputting and saving the point information. The present invention solves the problems in the prior art that the point condensation algorithm cannot distinguish the echo overlapping situation, is prone to large deviations between the true position of the target and the condensed position, and only one point trace appears for multiple batches of adjacent targets, and improves the efficiency and accuracy of track tracking in complex environments.

Description

Technical Field

[0001] The present invention relates to the technique of point clustering in radar signal processing, and particularly to a radar point clustering method based on echo envelope detection. Background Art

[0002] Radar target tracking technology is an important link in the data processing system. The input signal for radar target tracking is the points formed after the detection of the front-end radar signal processing. The radar receives the front-end echo signal and sends it to the signal processing to form discrete points that cross the threshold, namely EP points. In a two-coordinate pulse system radar, without considering the pitch, since the target occupies a certain size in both azimuth and range, and the radar also has a certain beam width and a certain range resolution, a batch of targets may be composed of multiple discrete or continuous EP points after detection and processing. In each radar scan cycle, all the EP points of a target are clustered into one point through the point clustering method, and the actual position information of the target is output as the information source for subsequent track initiation and target tracking. The generation of points requires a point clustering algorithm. In the prior art, the point clustering algorithms include the geometric center method, the mass center method, the power method, and the amplitude maximum method. However, these algorithms do not consider the situations that occur in complex environments such as multi-target crossing, parallelism, and a large amount of clutter residue, and cannot distinguish the echo overlapping situation, easily resulting in a large deviation between the true position of the target and the clustered position, and only one point is generated for multiple batches of adjacent targets, etc., increasing the difficulty of track tracking and having low processing efficiency and track accuracy. Summary of the Invention

[0003] Object of the Invention: The object of the present invention is to provide a radar point clustering method based on echo envelope detection that can distinguish the echo overlapping situation that occurs in complex environments such as multi-target crossing, parallelism, and a large amount of clutter residue, and has high processing efficiency and track accuracy.

[0004] Technical Solution: To achieve the above object, the radar point clustering method based on echo envelope detection described in the present invention includes the following steps:

[0005] Step S1: Receive the original radar detection point data and process it;

[0006] Step S2: Merge the detection points to form multiple large clusters of detection point information, and obtain the finally merged point ambiguity region;

[0007] Step S3: Determine whether the information in the point ambiguity region satisfies waveform resolution. If it satisfies, execute Step S4; if not, execute Step S6;

[0008] Step S4: Perform azimuth resolution processing on the point ambiguity region that satisfies waveform resolution to form one or more "small clusters" regions after resolution;

[0009] Step S5: Perform range resolution processing on the "small cluster" areas formed by azimuth resolution to form the final individual echo areas;

[0010] Step S6: Calculate the echo positions condensed in each area according to the amplitude weighting method for the echo areas, and output and save the echo information.

[0011] The step S1 of receiving and processing the original radar detection point data means: performing azimuth and range quantization on the radar echo data according to the radar pulse repetition period and the radar sampling rate, obtaining that there are L triggers in one radar scan circle, and there are N range cells on each trigger, that is, there are L×N sampling points in one radar scan circle, and the positions of the detection points that cross the threshold after signal processing on each trigger, namely the EP points, are where i2 represents the trigger serial number, j2 represents the EP point serial number, and mi2

[0012] is the total number of EP points on the i2-th trigger received and mi2≤N, the range cell is the range where the j2-th EP point on the i2-th trigger is located, and the azimuth is the azimuth where the j2-th EP point on the i2-th trigger is located.

[0013] The step S2 of merging the detection points to form multiple large cluster information of detection points and obtaining the finally merged echo ambiguity areas includes the following sub-steps:

[0014] Step S201: Form a line segment queue with detection points having the same range and different azimuths and the azimuth interval satisfying the following conditions:

[0015]

[0016] where, A0 is the azimuth where the current trigger is located, is the ending azimuth of line segment j3, and ΔDA min is the detection point azimuth break threshold;

[0017] Step S202: Merge all line segments whose azimuth interval and range interval simultaneously satisfy the following conditions into cluster information:

[0018]

[0019] where, ΔCA max is the azimuth break threshold between the cluster and the line segment, and ΔCR max is the range break threshold between the cluster and the line segment. The starting azimuth of cluster k is CAS k , 1≤k≤total number of clusters, the ending azimuth of cluster k is CAE k , and the starting range cell of cluster k is CRSk The end distance unit of cluster k is CRE k , is the starting azimuth of line segment j3, is the distance unit where line segment j3 is located, ΔCA k is the azimuth break between cluster k and line segment j3, ΔCR k is the distance break between cluster k and line segment j3;

[0020] ΔCA k and ΔCR k The calculation formulas are as follows:

[0021]

[0022] Step S203: Fuse all clusters whose azimuth intervals and distance intervals simultaneously meet the following conditions into large cluster information to obtain the finally merged point trace fuzzy region:

[0023]

[0024] Among them, ΔCCA k is the azimuth break between cluster k and cluster t, ΔCCR k is the distance break between cluster k and cluster t, CAW k is the azimuth width of cluster k, CAW t is the azimuth width of cluster t, CRW k is the distance width of cluster k, CRW t is the distance width of cluster t, k and t are the numbers of clusters;

[0025] Let the starting azimuth of cluster k be CAS k , 1 ≤ k ≤ total number of clusters, the ending azimuth is CAE k , the starting distance unit is CRS k , the ending distance unit is CRE k , the starting azimuth of the cluster t to be compared is CAS t , 1 ≤ t ≤ total number of clusters, and t ≠ k, the ending azimuth is CAE t , the starting distance unit is CRS t , the ending distance unit is CRE t , then the following calculation formulas hold:

[0026] ΔCCA k = |CAS k - CAE t | or ΔCCA k = |CAE k - CAS t |

[0027] ΔCCR k = |CRSk -CRE t | or ΔCCR k = | CRE k -CRS t |

[0028] CAW k = | CAE k -CAS k |

[0029] CAW t = | CAE t -CAS t |

[0030] CRW k = | CRE k -CRS k |

[0031] CRW t = | CRE t -CRS t |.

[0032] The determination in step S3 of whether the information within the plot ambiguity region meets waveform resolution means: According to the characteristics of a pulsed radar, affected by the radar beam width and pulse width, a target occupies a certain size in both azimuth and range. If the azimuth width and range width of the plot ambiguity region simultaneously meet the following conditions, then the plot azimuth resolution process in step S4 is performed on this region; if not, step S6 is executed.

[0033]

[0034] Among them, AW 点迹模糊区域 is the azimuth width of the plot ambiguity region, RW 点迹模糊区域 is the range width of the plot ambiguity region, θ 3dB is the 3dB beam width of the radar antenna, τ is the radar pulse width, and C is the speed of light.

[0035] The azimuth resolution process in step S4 for the plot ambiguity region that meets waveform resolution to form one or more "small cluster" regions after resolution means azimuth resolution of overlapping echoes based on echo envelope detection, including the following sub-steps:

[0036] Step S401: According to the pulsed radar system, the radar antenna pattern function is F(θ), where θ is the azimuth. The amplitude distribution of the radar target echo in the azimuth direction satisfies the standard waveform relationship of weak-strong-weak or valley-peak-valley. The azimuth echo envelope is formed by using the maximum amplitude values at different range points in the same azimuth within the fuzzy region block. When resolving the azimuth of the traces, the method of indexing local extrema with large differences in azimuth is adopted to find the positions of the peaks and valleys in the echo envelope. For the pulse compression radar system, affected by the target size and radar resolution, the target will occupy multiple range cells in the same azimuth. The EP points with the largest amplitude values in the same azimuth are selected to form the echo envelope;

[0037] Step S402: Search for all peak points and valley points in the azimuth echo envelope. Determine the positions of the peaks and valleys in the envelope by the extreme azimuth interval threshold I max and judge the correctness of the peaks and valleys by the peak-valley amplitude ratio threshold K max . When I max ≥ θ 3dB / 2, where θ 3dB is the 3dB beam width of the radar antenna, and K max ≥ the main-to-side lobe ratio. The specific methods for searching and judging the peaks and valleys are as follows:

[0038] Suppose there are n triggers in the echo envelope, and there are m1, m2, m3,..., mn EP detection points on each trigger respectively. The matrix queue formed by the amplitudes on each trigger is:

[0039]

[0040] The matrix queue formed by the range r and azimuth θ where the EP points on each corresponding trigger are located is:

[0041]

[0042] Select the EP points with the largest amplitude values in different range cells in the same azimuth to form the echo envelope max(A) = A' = [A'1 A'2 A'3... A' n , and the corresponding azimuth is θ' = [θ'1 θ'2 θ'3... θ' n ;

[0043] Take the i1-th amplitude value in the envelope and sequentially judge whether the following conditions are all satisfied. If they are satisfied, then it is the peak amplitude, and the azimuth of the corresponding EP point is the peak position θ 峰 ;

[0044]

[0045] Take the j1-th amplitude value in the envelope Successively determine whether the following conditions are simultaneously satisfied. If so, then is the trough amplitude, and the azimuth of the corresponding EP point is the trough position θ 谷

[0046]

[0047] If then the peak and the trough are valid extreme points; otherwise, they are invalid extreme points. Then, eliminate the invalid abnormal extreme points. After traversing, the number of peaks obtained is X ang , and the number of troughs is Y ang ;

[0048] Step S403: Split the echo region according to the standard waveform of valley-peak-valley to obtain one or more small region blocks after azimuth resolution, that is, sort all the extreme points in the azimuth according to the azimuth size to form an azimuth extreme value matrix where 0 ≤ x1 ≤ X ang , 0 ≤ y1 ≤ Y ang , and successively search for all the extreme points in the azimuth extreme value matrix that satisfy θ 谷 -θ 峰 -θ 谷 sequence. The corresponding azimuth region θ 谷 -θ 谷 forms the split azimuth region.

[0049] The step S5 performs range resolution processing on the "small cluster" region formed by azimuth resolution to form the final individual plot regions, which means: Range resolution of overlapping echoes based on echo envelope detection includes the following sub-steps:

[0050] Step S501: For a pulse compression radar system, when the transmitted signal is a linear frequency modulation signal, after the received signal is pulse-compressed by a matched filter, the envelope of the formed echo signal is approximately a sinc function. The amplitude distribution of the radar target echo in the range satisfies the standard waveform relationship of weak-strong-weak or valley-peak-valley. Perform range resolution processing on each small region block after azimuth resolution respectively, and form a range echo envelope by taking the maximum amplitude value at different azimuth range points within the same range in the small region, that is, select the EP points with the largest amplitude value at different azimuths in the same range cell to form a range echo envelope;

[0051] Step S502: Search for all the peak points and trough points in the range echo envelope, and determine the positions of the peaks and troughs in the envelope by the extreme range interval threshold J max , and the peak and trough amplitude ratio threshold L maxJudge the correctness of the wave peaks and wave troughs, J max The value of L is related to factors such as the pulse width, radar resolution, and target size max ≥ the main-to-side lobe ratio. The specific methods for searching and judging wave peaks and wave troughs are as follows:

[0052] Suppose there are q triggers in a certain area after azimuth resolution, and there are p1, p2, p3,..., pq EP detection points on each trigger respectively. The matrix queue composed of the amplitudes on each trigger is:

[0053]

[0054] The distance r where the EP points are located on each corresponding trigger * and the azimuth θ * The matrix queue formed is:

[0055]

[0056] Select the EP points with the largest amplitude values at the same distance but different azimuths to form the echo envelope where p max is the maximum value of the distance points, and the corresponding distance point is

[0057] Take the i3-th amplitude value in the envelope Judge in turn whether the following conditions are simultaneously satisfied. If so, then B' i is the wave peak amplitude, and the distance of the corresponding EP point is the wave peak position r 峰 ,

[0058]

[0059] Take the j4-th amplitude value in the envelope Judge in turn whether the following conditions are simultaneously satisfied. If so, then is the wave trough amplitude, and the distance of the corresponding EP point is the wave trough position r 谷 ,

[0060]

[0061] If then the wave peak and the wave trough are valid extreme points, otherwise they are invalid extreme points. Then eliminate the invalid abnormal extreme points. After traversing, the number of wave peaks is X rng , and the number of wave troughs is Y rng ;

[0062] Step S503: Split the echo region according to the standard waveform of valley-peak-valley to obtain one or more small regional blocks after range resolution, that is, sort all the extreme points on the range envelope according to the range size to form a range extreme value matrix E rng =[r 谷x r 峰y , where 0 ≤ x2 ≤ X rng , 0 ≤ y2 ≤ Y rng , sequentially search for all extreme points in the range extreme value matrix that satisfy r 谷 -r 峰 -r 谷 order, and the corresponding range region r 谷 -r 谷 forms the split range region.

[0063] In the step S6 described above, the dot position condensed in each region is calculated according to the amplitude weighting method, and the dot information is output and saved. The specific calculation method is as follows:

[0064] Assume that there are M EP points in a certain region, and the range and azimuth coordinates of each EP point are p1 = (r1, θ1), p2 = (r2, θ2), p3 = (r3, θ3), ……, p M =(r M , θ M ), and the amplitude values of each EP point are A1, A2, A3, ……, A M , then the range r center and azimuth angle θ center of the condensed point calculated by the amplitude weighting method are as follows:

[0065]

[0066] Beneficial effects: The radar dot condensation method based on echo envelope detection of the present invention, on the basis of obtaining a fuzzy dot region through dot merging, then obtains each refined dot region through dot azimuth and range resolution, and finally uses the amplitude weighting method to complete the calculation of the dot center position in each region, completing the dot condensation process completely. It has the following advantages: 1. A complete and effective splitting solution is provided for the situation where targets and targets, targets and clutter echoes overlap, resulting in incorrect dotting or inaccurate dot positions, improving the tracking accuracy in complex scenarios such as target encounters, intersections, parallelism, and clutter penetration in subsequent radar tracking systems; 2. Only special processing is performed for the situation of echo overlap, the overall scheme has a small increased computational load, the calculation process is simple, can be processed in real time, and has low requirements for hardware performance, and does not increase the cost of hardware equipment while realizing the functions of the present invention. Description of the Drawings

[0067] Figure 1 Flow chart of the radar point track condensation method based on echo envelope detection according to the present invention;

[0068] Figure 2 Simulation result of resolving overlapping point tracks of targets and targets;

[0069] Figure 3 Simulation result of resolving overlapping point tracks of targets and clutter;

[0070] Figure 4 Schematic diagram of improving the continuity of target tracking in the clutter area by point track resolution in a certain radar tracking system according to the present invention. Specific implementation mode

[0071] The technical solution of the present invention will be described in detail below in conjunction with the embodiments and the drawings.

[0072] As Figure 1 shown, the radar point track condensation method based on echo envelope detection according to the present invention includes the following steps:

[0073] Step S1: During the point track merging process, with the EP point after signal processing as the center, search for its distance unit relationship and azimuth difference with the existing line segments, and establish or update the line segment queue with the same distance unit and different azimuths. The condition for the fusion of the EP point and the existing line segment is that the distance units are equal, the azimuths have an intersection or the difference is less than the break threshold. Among them, the azimuth break threshold is related to the number of triggers in one circle of the radar, that is, the radar period and the radar pulse repetition frequency. The more triggers, the smaller the azimuth break threshold, and the larger the trigger break threshold; the fewer triggers, the larger the azimuth break threshold, and the smaller the trigger break threshold. The azimuth break threshold is set between 0.44° and 0.88°. In the present invention, the judgment formula for establishing or updating the line segment is as follows:

[0074]

[0075] wherein, A0 is the azimuth where the current trigger is located, is the end azimuth of line segment j3, and ΔDA min is the azimuth break threshold of the detection point.

[0076] Step S2: Fuse all line segments that simultaneously meet the following conditions for the azimuth interval and the distance interval into cluster information:

[0077]

[0078] wherein,

[0079]

[0080] ΔCA max is the azimuth break threshold between the cluster and the line segment, and ΔCR maxis the distance break threshold between the cluster and the line segment. The starting azimuth of cluster k (1 ≤ k ≤ total number of clusters) is CAS k , and the ending azimuth is CAE k , the starting distance unit is CRS k , and the ending distance unit is CRE k , ΔCA k is the azimuth break between cluster k and line segment j, and ΔCR k is the distance break between cluster k and line segment j is the starting azimuth of line segment j3 is the distance unit where line segment j3 is located.

[0081] Then, all clusters whose azimuth intervals and distance intervals simultaneously meet the following conditions are merged into large cluster information to obtain the final merged target fuzzy region:

[0082]

[0083] Among them, ΔCCA k is the azimuth break between cluster k and cluster t, and ΔCCR k is the distance break between cluster k and cluster t. CAW k is the azimuth width of cluster k, CAW t is the azimuth width of cluster t, CRW k is the distance width of cluster k, CRW t is the distance width of cluster t, where k and t are the numbers of clusters:

[0084] Let the starting azimuth of cluster k (1 ≤ k ≤ total number of clusters) be CAS k , and the ending azimuth be CAE k , the starting distance unit be CRS k , and the ending distance unit be CRE k , and the starting azimuth of the compared cluster t (1 ≤ t ≤ total number of clusters, and t ≠ k) be CAS t , and the ending azimuth be CAE t , the starting distance unit be CRS t , and the ending distance unit be CRE t , then there are the following calculation formulas:

[0085] ΔCCA k =|CAS k -CAE t | or ΔCCA k =|CAE k -CAS t |

[0086] ΔCCR k =|CRS k -CRE t| or ΔCCR k = | CRE k - CRS t |

[0087] CAW k = | CAE k - CAS k |

[0088] CAW t = | CAE t - CAS t |

[0089] CRW k = | CRE k - CRS k |

[0090] CRW t = | CRE t - CRS t |。

[0091] Step S3: According to the pulsed radar system and the radar antenna pattern function F(θ), where θ is the azimuth, it is known that the amplitude distribution of the radar target echo in the azimuth direction satisfies the standard waveform relationship of weak - strong - weak or valley - peak - valley. Store all the EP point information in the large cluster information formed after merging, and then perform matching operations using the standard waveform. Define the peak points and valley points that satisfy the valley - peak - valley waveform as extreme points. If more than two sets of extreme points that satisfy the valley - peak - valley distribution are found, then it is considered that this large cluster of echoes is composed of more than one target, and perform point - trace azimuth resolution processing. If the above - mentioned extreme points that satisfy the valley - peak - valley distribution are not found, then perform amplitude weighting processing on the ambiguous point - trace area to obtain the point - trace information condensed in each area, and output and save it.

[0092] Step S4: Traverse all the large clusters of detection points, and perform azimuth resolution processing on the large cluster information that meets the conditions such as azimuth width and distance width according to the echo envelope detection method to form the "small cluster" information after splitting. In the present invention, the simulation effects of target - target overlapping point - trace resolution and target - clutter overlapping point - trace resolution are as Figure 2 shown, and the specific process is as follows:

[0093] The criterion for extracting the azimuth echo envelope is:

[0094] Assume that there are n triggers in the echo envelope, and there are m1, m2, m3,..., mn EP detection points on each trigger respectively. The matrix queue composed of the amplitudes on each trigger is:

[0095]

[0096] The matrix queue composed of the distance r and azimuth θ of the EP points corresponding to each trigger is as follows:

[0097]

[0098] Select the EP points with the largest amplitude value at different distances in the same azimuth to form the echo envelope max(A)=A'=[A'1 A'2 A'3... A' n , and the corresponding azimuth is θ'=[θ'1 θ'2 θ'3... θ' n .

[0099] The method for finding the peak points and valley points in the azimuth echo envelope is as follows:

[0100] Let I max be the extreme azimuth interval threshold, and K max be the peak and valley amplitude ratio threshold, I max ≥θ 3dB / 2, K max ≥ main-to-side lobe ratio, take the i1-th amplitude value in the envelope Judge in turn whether the following conditions are simultaneously satisfied. If so, then is the peak amplitude, and the azimuth of the corresponding EP point is the peak position θ 峰 .

[0101]

[0102] Take the j1-th amplitude value in the envelope Judge in turn whether the following conditions are simultaneously satisfied. If so, then is the valley amplitude, and the azimuth of the corresponding EP point is the valley position θ 谷 .

[0103]

[0104] If then the peak and the valley are valid extreme points, otherwise they are invalid extreme points. Then eliminate the invalid abnormal extreme points. After traversing, the number of peak points is X ang , and the number of valley points is Y ang .

[0105] The criterion for determining the azimuth resolution region of the traces is as follows:

[0106] Sort all the extreme points in the azimuth according to the azimuth size to form the azimuth extreme value matrix where 0≤x1≤X ang , 0≤y1≤Y ang, sequentially search for all extreme points in the azimuth extreme value matrix that satisfy θ 谷 -θ 峰 -θ 谷 The corresponding azimuth regions θ 谷 -θ 谷 constitute the split azimuth regions.

[0107] Step S5: Traverse each small region block after azimuth resolution, and perform range resolution processing on each small region block respectively. In the present invention, the simulation effect of target and target aliased echo point resolution and the simulation effect of target and clutter aliased echo point resolution are as Figure 3 shown, and the specific process is as follows:

[0108] The criterion for range echo envelope extraction is:

[0109] Assume that there are q triggers in a certain region after azimuth resolution, and there are p1, p2, p3,..., pq EP detection points on each trigger respectively. The matrix queue composed of the amplitudes on each trigger is:

[0110]

[0111] The corresponding range r where the EP points are located on each trigger * and azimuth θ * constitute the matrix queue:

[0112]

[0113] Select the EP points with the largest amplitude values at the same range and different azimuths to form the range echo envelope The corresponding range point is where p max is the maximum value of the range point.

[0114] The method for finding the peak points and valley points in the range echo envelope is:

[0115] Let J max be the extreme range interval threshold, and L max be the peak and valley amplitude ratio threshold. The value of J max is related to the pulse width, radar resolution, target size, etc., and should be determined according to the actual engineering situation. L max ≥ the range main lobe to sidelobe ratio, take the i3th amplitude value in the envelope Sequentially judge whether the following conditions are simultaneously satisfied. If satisfied, then B i ' is the peak amplitude, and the corresponding EP point range is the peak position r 峰 .

[0116]

[0117] Take the j4th amplitude value in the envelope Judge in turn whether the following conditions are satisfied simultaneously. If so, then is the trough amplitude, and the corresponding distance of the EP point is the trough position r 谷 .

[0118]

[0119] If then the peak and the trough are valid extreme points, otherwise they are invalid extreme points. Then eliminate the invalid abnormal extreme points. After traversing, the number of peaks obtained is X rng , and the number of troughs is Y rng .

[0120] The criterion for determining the point trace distance resolution region is as follows:

[0121] Sort all the extreme points in terms of distance to form a distance extreme value matrix E rng =[r 谷x r 峰y , where 0 ≤ x2 ≤ X rng , 0 ≤ y2 ≤ Y rng , and sequentially search for all the extreme points in the distance extreme value matrix that satisfy r 谷 -r 峰 -r 谷 sequence. The corresponding distance region r 谷 -r 谷 forms the split distance region.

[0122] Step S6: Calculate the position information of the point traces for each region after azimuth and distance resolution respectively using the amplitude weighting method. The specific process is as follows:

[0123] Assume that there are M EP points in a certain detection point region after point trace resolution. The distance and azimuth coordinates of each EP point are p1 = (r1, θ1), p2 = (r2, θ2), p3 = (r3, θ3), ……, p M =(r M ,θ M ), and the amplitude values of each EP point are A1, A2, A3, ……, A M . Then the distance r center and azimuth angle θ center of the condensation point calculated using the amplitude weighting method are calculated as follows:

[0124]

[0125] If Figure 4As shown in the figure, it is a screenshot of the radar display during a certain radar tracking process. The batch numbers of the two tracking targets are TV0104 and TV0209 respectively. During the tracking process, the two targets encounter and pass through clutter situations, and the echoes of the targets and the clutter overlap. Moreover, the overlap period and the clutter-passing time are relatively long. By using the radar point clustering method based on echo envelope detection proposed by the present invention, the true target's point position can be correctly found when the echoes overlap, enabling both targets to be tracked normally and stably without batch mixing or batch changing, verifying the effectiveness and practicality of this method.

Claims

1. A radar point track condensation method based on echo envelope detection, characterized in that: It includes the following steps: Step S1: Receive the original radar detection point data and process it; Step S2: Merge the detection points to form multiple large clusters of detection point information, and obtain the finally merged point track ambiguity region; Step S3: Determine whether the information in the point track ambiguity region satisfies waveform resolution. If it does, execute Step S4; if not, execute Step S6; Step S4: Perform azimuth resolution processing on the point track ambiguity region that satisfies waveform resolution to form one or more "small cluster" regions after resolution; Step S5: Perform range resolution processing on the "small cluster" regions formed by azimuth resolution to form the final individual point track regions; Step S6: Calculate the point track positions condensed in each region according to the amplitude weighting method for the point track regions, and output and save the point track information; When Step S3 determines whether the information in the point track ambiguity region satisfies waveform resolution, it means that: according to the characteristics of the pulsed radar system, affected by the radar beam width and pulse width, the target occupies a certain size in both azimuth and range. If the azimuth width and range width of the point track ambiguity region simultaneously meet the following conditions, then the point track azimuth resolution processing in Step S4 is performed on this region; if not, Step S6 is executed; Among them, AW 点迹模糊区域 is the azimuth width of the plot ambiguity region, RW 点迹模糊区域 is the range width of the plot ambiguity region, θ 3dB is the 3dB beam width of the radar antenna, τ is the radar pulse width, and C is the speed of light; When Step S4 performs azimuth resolution processing on the point track ambiguity region that satisfies waveform resolution to form one or more "small cluster" regions after resolution, it means performing azimuth resolution on the overlapping echoes based on echo envelope detection, including the following sub-steps: Step S401: According to the pulsed radar system, the radar antenna pattern function F(θ), where θ is the azimuth, the amplitude distribution of the radar target echo in the azimuth satisfies the standard waveform relationship of weak-strong-weak or valley-peak-valley. Use the maximum amplitude values at different ranges but the same azimuth within the ambiguity region block to form the azimuth echo envelope. When performing point track azimuth resolution, use the method of indexing local extrema with large differences in azimuth to find the positions of the peaks and valleys in the echo envelope. For the pulse compression radar system, affected by the target size and radar resolution, the target occupies multiple range cells at the same azimuth. Select the EP points with the largest amplitude value at the same azimuth to form the echo envelope; Step S402: Search for all the peak points and valley points in the azimuth echo envelope. Determine the positions of the peaks and valleys in the envelope according to the extreme azimuth interval threshold I max and the peak-valley amplitude ratio threshold K max to judge the correctness of the peaks and valleys. If I max ≥θ 3dB / 2, where θ 3dB is the 3dB beam width of the radar antenna, and K max ≥ the main-to-side lobe ratio, the specific methods for searching and judging the peaks and valleys are as follows: Assume that there are n triggers in the echo envelope, and there are m1, m2, m3,..., mn EP detection points on each trigger respectively. The matrix queue formed by the amplitudes on each trigger is: The matrix queue formed by the range r and azimuth θ where the EP points are located on each corresponding trigger is: Select the EP points with the largest amplitude values on different range cells in the same azimuth to form the echo envelope max(A)=A'=[A′1 A′2 A′3... A′ n , and the corresponding azimuth is θ'=[θ′1 θ′2 θ′3... θ′ n ; Take the i1-th amplitude value in the envelope Judge in sequence whether the following conditions are simultaneously satisfied. If so, then is the peak amplitude, and the azimuth of the corresponding EP point is the peak position θ 峰 ; Take the j1-th amplitude value in the envelope Successively determine whether the following conditions are simultaneously satisfied. If so, then is the trough amplitude, and the azimuth of the corresponding EP point is the trough position θ 谷 If then the wave crest and the wave trough are valid extreme points, otherwise they are invalid extreme points. Then, the invalid abnormal extreme points are removed, and the number of wave crests obtained after traversal is X ang , and the number of wave troughs is Y ang ; Step S403: Split the echo region according to the standard waveform of valley-peak-valley to obtain one or more small regional blocks after azimuth resolution, that is, sort all the extreme points in the azimuth according to the azimuth size to form an azimuth extreme value matrix where 0 ≤ x1 ≤ X ang , 0 ≤ y1 ≤ Y ang , sequentially search for all the extreme points in the azimuth extreme value matrix that satisfy θ 谷 -θ 峰 -θ 谷 sequence, and the corresponding azimuth region θ 谷 -θ 谷 constitute the split azimuth region.

2. The radar point track condensation method based on echo envelope detection according to claim 1, characterized in that: The step S1 of receiving and processing the original radar detection point data means: quantizing the azimuth and distance of the radar echo data according to the radar pulse repetition period and the radar sampling rate, obtaining that there are L triggers in one radar scan circle, and there are N range cells on each trigger, that is, there are L×N sampling points in one radar scan circle, and the position of each detection point that passes the threshold and is formed after signal processing on each trigger, namely the EP point, is where i2 represents the trigger serial number, j2 represents the EP point serial number, mi2 is the total number of EP points on the i2-th trigger received and mi2≤N, range cell is the range where the j2-th EP point on the i2-th trigger is located, azimuth is the azimuth where the j2-th EP point on the i2-th trigger is located.

3. The radar point track condensation method based on echo envelope detection according to claim 1, characterized in that: The sub-steps included in Step S2 for merging the detection points to form multiple large clusters of detection point information and obtaining the finally merged point track ambiguity region are as follows: Step S201: Form a line segment queue with detection points that have the same range, different azimuths, and the azimuth interval satisfies the following conditions: where A0 is the azimuth where the current trigger is located, is the end azimuth of line segment j3, and ΔDA min is the azimuth break threshold of the detection point; Step S202: Merge the line segments whose azimuth intervals and range intervals simultaneously satisfy the following conditions into cluster information: where, ΔCA max is the azimuth break threshold between the cluster and the line segment, ΔCR max is the distance break threshold between the cluster and the line segment, the starting azimuth of cluster k is CAS k , 1 ≤ k ≤ total number of clusters, the ending azimuth of cluster k is CAE k , the starting distance unit of cluster k is CRS k , the ending distance unit of cluster k is CRE k , is the starting azimuth of line segment j3, is the distance unit where line segment j3 is located, ΔCA k is the azimuth break between cluster k and line segment j3, ΔCR k is the distance break between cluster k and line segment j3; ΔCA k and ΔCR k are calculated as follows: Step S203: Merge all the clusters whose azimuth intervals and range intervals simultaneously satisfy the following conditions into large cluster information to obtain the finally merged point track ambiguity region: Among them, ΔCCA k is the azimuth fracture between cluster k and cluster t, ΔCCR k is the distance fracture between cluster k and cluster t, CAW k is the azimuth width of cluster k, CAW t is the azimuth width of cluster t, CRW k is the distance width of cluster k, CRW t is the distance width of cluster t, and k and t are the numbers of clusters; Let the starting orientation of cluster k be CAS k , where 1 ≤ k ≤ total number of clusters, and the ending orientation be CAE k , the starting distance unit be CRS k , the ending distance unit be CRE k , the starting orientation of the compared cluster t be CAS t , where 1 ≤ t ≤ total number of clusters and t ≠ k, and the ending orientation be CAE t , the starting distance unit be CRS t , the ending distance unit be CRE t , then the following calculation formula holds: ΔCCA k = |CAS k - CAE t | or ΔCCA k = |CAE k - CAS t | ΔCCR k = |CRS k - CRE t | or ΔCCR k = |CRE k - CRS t | CAW k = |CAE k - CAS k | CAW t = |CAE t - CAS t | CRW k = |CRE k - CRS k | CRW t = |CRE t - CRS t |.

4. The method for radar echo clustering based on echo envelope detection according to claim 1, wherein: The step S5 performs range resolution processing on the "small cluster" area formed by azimuth resolution, and the formation of the final individual echo area means that range resolution of overlapping echoes based on echo envelope detection includes the following sub-steps: Step S501: For a pulse compression radar system, when the transmitted signal is a linear frequency modulation signal, after the received signal is pulse-compressed by a matched filter, the envelope of the formed echo signal is approximately a sinc function. The amplitude distribution of the radar target echo in range satisfies the standard waveform relationship of weak-strong-weak or valley-peak-valley. Range resolution processing is performed on each small area block after azimuth resolution. The maximum amplitude values at the same range but different azimuth-range points within the small area are combined to form a range echo envelope, that is, the EP points with the maximum amplitude values at the same range cell but different azimuths are selected to form the range echo envelope; Step S502: Search for all the peak points and valley points in the echo envelope, and determine the positions of the peaks and valleys in the envelope by the extreme distance interval threshold J max and the peak and valley amplitude ratio threshold L max to judge the correctness of the peaks and valleys. The value of J max is related to factors such as the pulse width, radar resolution, and target size. L max ≥ the main-to-side lobe ratio. The specific methods for searching and judging peaks and valleys are as follows: Suppose there are q triggers in a certain area after azimuth resolution, and there are p1, p2, p3,..., pq EP detection points on each trigger respectively. The matrix queue formed by the amplitudes on each trigger is: The distance r where the EP point is located on each corresponding trigger * and the azimuth θ * The matrix queue formed is as follows: Select the EP points with the largest amplitude values at the same distance but different azimuths to form the echo envelope where p max is the maximum value of the distance points, and the corresponding distance point is Take the i3-th amplitude value in the envelope Successively determine whether the following conditions are simultaneously satisfied. If so, then B i ' is the peak amplitude, and the corresponding distance of the EP point is the peak position r 峰 , Take the j4-th amplitude value in the envelope Successively determine whether the following conditions are simultaneously satisfied. If so, then is the trough amplitude, and the distance of the corresponding EP point is the trough position r 谷 , If then the wave crest and the wave trough are valid extreme points, otherwise they are invalid extreme points. Then, the invalid abnormal extreme points are removed, and the number of wave crests obtained after traversal is X rng , and the number of wave troughs is Y rng ; Step S503: Split the echo region according to the standard waveform of valley-peak-valley to obtain one or more small region blocks after range resolution, that is, sort all the extreme points on the range envelope according to the range size to form a range extreme value matrix E rng = [r 谷x r 峰y , where 0 ≤ x2 ≤ X rng , 0 ≤ y2 ≤ Y rng , sequentially search for all extreme points in the range extreme value matrix that satisfy r 谷 -r 峰 -r 谷 order, and the corresponding range region r 谷 -r 谷 forms the split range region.

Images