A track initiation method based on time-sharing blind spot filling strategy and multi-level information matching
By using time-sharing blinding strategy and multi-level information matching in the start of the track, the correlation between short-period pulses and Doppler speed are solved by solving the large amount of calculation and noise removal difficulties of traditional methods in the context of strong clutter. Problem, achieving a faster and more accurate track start.
Patent Information
- Application Number
- CN202210888182.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-07-27
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2042-07-27
AI Technical Summary
The traditional track start method has a large amount of calculation in the context of strong clutter and is difficult to effectively remove noise, resulting in a longer start time and reduced accuracy, and difficulty in selecting thresholds lead to track splitting.
Using a method based on the time-sharing blind-compensation strategy and multi-level information matching, the correlation between short-period pulses and Doppler speed are calculated by processing the data after the pulse pressure is calculated, and the distance division point of long-period pulses is defused. Match with the point track to determine whether the target can be started and ensure the accuracy and speed of the track starting.
Improve the speed and accuracy of track start, reduce the start time, enhance the fast automatic start capability for high-speed targets, and improve the criterion of track start through multi-level information matching, meeting tactical needs .
Smart Images

Figure CN115166718B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of track initiation in radar data processing. Background Art
[0002] Multi-target tracking is one of the most basic tasks of modern radar. Track initiation is an important function in multi-target tracking technology, which can automatically establish tracks for targets entering the radar power area. Track initiation algorithms mainly include: intuitive method, logical method and Hough transform method. Among them, the intuitive method and logical method are suitable for track initiation under relatively weak clutter background, with small calculation amount and high execution efficiency; while the Hough transform method and its improved algorithm are more suitable for track initiation under strong clutter background, but the calculation amount is large.
[0003] Traditional track initiation methods are all based on point-level, and the main steps are: establishing the original track, establishing a temporary track according to the initial association gate, and confirming the temporary track according to the predicted association gate to form a stable track. The original track is often accompanied by false track points such as noise and interference. Therefore, in the process of forming a stable track, the problem of threshold selection should be considered. If the threshold value is too small, the start time will increase; if the threshold value is too large, it will cause target error association, and the same target will have multiple stable tracks, that is, track splitting will occur.
[0004] If useful information can be extracted from the wave position level data for track initiation, it can not only reduce the initiation time, but also improve the initiation accuracy. At present, radars mostly use pulse accumulation to improve the signal-to-noise ratio, or process moving targets to remove clutter, etc., to provide data guarantee for target tracking or initiation. The processing between multiple pulses can obtain the Doppler information of the target for speed measurement. At the same time, due to the residual energy of the high repetition rate pulse, there will be targets that exceed the maximum range of the current pulse. By resolving the distance ambiguity, the target distance can be obtained. Although these are traditional methods of ranging radar, if the correlation between multiple PRI pulses is used, not only can the interference noise be removed, but also the useful information of the target point trace can be extracted from the wave position level data for track initiation. Summary of the invention
[0005] In order to solve the above technical problems, the present invention proposes a track initiation method based on a time-sharing blind spot filling strategy and multi-level information matching, and the technical solution includes:
[0006] (1) Process the data after pulse pressure modulus calculation, calculate the folding times M for each distance unit according to the correlation between short-period pulses, and mark them. The distance unit parameters are statistically calculated in units of pulses as shown in equations (1) and (2);
[0007] δ′ i ={δ 1, ,δ2,…,δ m} (1)
[0008] δ j ={a δ ,u,r δ ,c,t,v,f}; (2)
[0009] Among them, a δ 、u、r δ , c, t, v and f are the azimuth, distance unit, defuzzified point distance, folding times, time, speed and mark of the position that meets the criteria, i = 1, 2, ..., n, n is the number of short pulses in the pulse group, j = 1, 2, ..., m, m is the number of distance units in the pulse that meets the criteria;
[0010] (2) According to the correlation between short-period pulses, the Doppler velocity of the corresponding distance unit is calculated, as shown in formula (3):
[0011] v di = {v d1, ,v d2 ,…,v dm} (3)
[0012] And record it in the corresponding distance unit vector δ j In the velocity attribute, j is the index of the distance unit that meets the criteria within the pulse;
[0013] (3) Pulse group accumulation, constant false alarm processing and point trace detection aggregation are performed. At this time, a δ and u δ are the direction and distance of the point trace after condensation;
[0014] (4) According to the distance dividing point between the long and short cycles, the point traces of the markers in the short cycle pulse distance segment are defuzzified, as shown in equations (4) and (5):
[0015] t r =MT+t (4)
[0016] R=(c*t r ) / 2 (5)
[0017] Among them, T is the short wave pulse repetition period, t is the short pulse delay, and the solved R is stored in the vector δ j r δ ;
[0018] (5) At the same time, the point trace record within the long and short period pulse distance segment is shown in formula (6):
[0019] γ k ={a γ ,r γ ,t γ}; (6)
[0020] Where k = 1, 2, ..., L, L is the number of points in the long pulse distance segment, in order to confirm the calculated target distance r δ is the folding target, the direction is a δ The distance is traversed from the boundary point of long and short pulses to the maximum range, and the point traces in the long-period pulse distance segment and the calculated target point traces meet the optimal solution d of the Euclidean distance criterion. min , as shown in formula (7):
[0021]
[0022] Among them, d min =min(d), record the vector δ corresponding to the optimal solution at this time j , and set the discrimination flag f = 1;
[0023] (6) The point trace vector δ j The point with the discriminant flag f = 1 is used as the candidate point for the start of the track and recorded in the starting point vector, as shown in formula (8):
[0024]
[0025] Where l = 1, 2, ..., l max , l max is the number of traces that can be started within the radar cycle, a, d, e, t are the azimuth, distance, elevation and time of the current trace respectively;
[0026] (7) Repeat steps (1) to (6) to calculate the Doppler velocity component v for the solved point trace. dx and v dy , and the unique component and As shown in equations (9), (10), (11) and (12):
[0027] v dx =v d *cos(a δ ) (9)
[0028] v dy =v d *sin(a δ ) (10)
[0029]
[0030]
[0031] Among them, T ris the radar period; Substitute the displacement component into the formula ρ=xcos(θ)+ysin(θ) to calculate the Doppler velocity displacement corresponding to According to the track start criteria, determine whether the inter-cycle displacement calculated by the Doppler velocity of the solution point track is consistent with the start vector If the point track in matches, the judgment flag f=2 will be set to allow the track to start.
[0032] The present invention realizes multi-level information matching of wave position level and point track level. Firstly, a pair of time-sharing blind wave positions of long-short period combination are set in the key focus area to ensure full-time coverage of the target. The energy surplus of short-period high-repetition frequency pulses is used to resolve the distance ambiguity of the echo data in the effective distance segment, and a long-period wave position is used to match the resolved data. Secondly, the correlation between the short-period high-repetition frequency pulses is used to calculate the Doppler velocity corresponding to each distance unit in the short-period wave position. Then, the displacement of the target adjacent antenna period is obtained by using the continuity of the target movement. Finally, the matching relationship between the Doppler velocity and the displacement is used to make a judgment to determine whether the target can start the track.
[0033] The present invention integrates the processing method of the wave position level into the processing of the track initiation. According to the detection requirements of the radar, in order to meet the requirements of the warning area for rapid automatic initiation of high-speed targets, a time-sharing blind wave position is designed (with high repetition rate pulses in the wave position), and the data after the pulse pressure modulus is solved for distance ambiguity. Before the point track is formed, the distance unit corresponding to the potential starting point track is marked (the track initiation is after the point track is formed), and data matching is achieved at the point track level. The present invention greatly improves the speed of track initiation; on the other hand, it increases the criterion for track initiation and improves the accuracy of initiation; more importantly, according to tactical requirements, the time-sharing blind wave position can be set to meet the detection requirements of key areas of concern.
[0034] The present invention is further described in detail below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1 Algorithm flow chart.
[0036] Figure 2 Fill in the blind positions in time.
[0037] Figure 3 Schematic diagram of point matching between time-sharing blind pulse groups.
[0038] Figure 4 Schematic diagram of the starting point of the track. DETAILED DESCRIPTION
[0039] The present invention proposes a track initiation method based on a time-sharing blind spot filling strategy and multi-level information matching. First, for the distance area of key concern, a pair of time-sharing blind spot filling wave positions of long and short period combination are set to ensure full-time coverage in the concerned area. Figure 2 As shown. Among them, T S1 The launch time of wave position 1, T R1 is the receiving time of wave position 1, T S2 The launch time of wave position 2, T R2 is the receiving time of wave position 2, and R is the effective distance covered by the time-sharing blind wave position. At the same time, the energy surplus of the high repetition rate pulse is used to resolve the distance ambiguity of the echo data within its effective distance segment. For the data after the distance ambiguity is resolved, it is necessary to determine whether the target of the distance unit is determined to be a folded target outside the maximum range of the wave position. Therefore, a long-period wave position with energy covering the blind wave is required, such as Figure 3 The middle wave position 3 is shown as shown, and according to the criteria, it is judged whether the target (dashed line target) after defuzzification in wave position 1 is the real target within the range of wave position 3. Then, the correlation between high repetition rate pulses is used to calculate the Doppler velocity of the corresponding position of each distance in the short wave position, and the displacement of the target in the adjacent cycle in the key distance area is obtained according to the starting criterion of Hough transform by using the correlation between cycles, and combined with the Doppler velocity, it is judged whether the target can be started. Figure 4 As shown in the real track marked in the figure, the displacement of the target itself conforms to the law of motion based on its Doppler velocity. On the contrary, if it is clutter, its displacement is random and irregular, and its Doppler velocity is often far below the threshold value.
[0040] The steps of the embodiment are as follows:
[0041] (1) Process the data after pulse pressure modulus calculation, calculate the folding times M for each distance unit according to the correlation between short-period pulses, and mark them. The distance unit parameters are statistically calculated in units of pulses as shown in equations (13) and (14);
[0042] δ′ i ={δ 1, ,δ2,…,δ m}; (13)
[0043] δ j ={a δ ,u,r δ ,c,t,v,f}; (14)
[0044] Among them, a δ 、u、r δ, c, t, v and f are the azimuth, distance unit, defuzzified point distance, folding times, time, speed and mark of the position that meets the criteria, i = 1, 2, ..., n, n is the number of short pulses in the pulse group, j = 1, 2, ..., m, m is the number of distance units in the pulse that meets the criteria;
[0045] (2) According to the correlation between short-period pulses, the Doppler velocity of the corresponding range unit is calculated, as shown in formula (15):
[0046] v di = {v d1, ,v d2 ,…,v dm} (15)
[0047] And record it in the corresponding distance unit vector δ j In the velocity attribute, j is the index of the distance unit that meets the criteria within the pulse;
[0048] (3) Pulse group accumulation, constant false alarm processing and point trace detection aggregation are performed. At this time, a δ and u δ are the direction and distance of the point trace after condensation;
[0049] (4) According to the distance dividing point between the long and short cycles, the point traces of the markers in the short cycle pulse distance segment are defuzzified, as shown in equations (16) and (17):
[0050] t r =MT+t (16)
[0051] R=(c*t r ) / 2 (17)
[0052] Among them, T is the short wave pulse repetition period, t is the short pulse delay, and the solved R is stored in the vector δ j r δ .
[0053] (5) At the same time, the point trace record within the long and short period pulse distance segment is shown in formula (18):
[0054] γ k ={a γ ,r γ ,t γ}; (18)
[0055] k=1,2,…,L, L is the number of points in the long pulse distance segment, in order to confirm the calculated target distance r δ is the folding target, the direction is a δThe distance is traversed from the boundary point of long and short pulses to the maximum range, and the point traces in the long-period pulse distance segment and the calculated target point traces meet the optimal solution d of the Euclidean distance criterion. min , as shown in formula (19):
[0056]
[0057] Among them, d min =min(d), record the vector δ corresponding to the optimal solution at this time j , and set the discrimination flag f = 1;
[0058] (6) The point trace vector δ j The point with the discriminant flag f = 1 is used as the candidate point for the start of the track and recorded in the starting point vector, as shown in formula (20):
[0059]
[0060] Where l = 1, 2, ..., l max , l max is the number of traces that can be started within the radar cycle, a, d, e, t are the azimuth, distance, elevation and time of the current trace respectively;
[0061] (7) Repeat steps (1) to (6) to calculate the Doppler velocity component v for the solved point trace. dx and v dy , and the unique component and As shown in equations (21), (22), (23) and (24):
[0062] v dx =v d *cos(a δ ) (twenty one)
[0063] v dy =v d *sin(a δ ) (twenty two)
[0064]
[0065]
[0066] Among them, T r is the radar period; Substitute the displacement component into the formula ρ=xcos(θ)+ysin(θ) to calculate the Doppler velocity displacement corresponding to According to the track start criteria, determine whether the inter-cycle displacement calculated by the Doppler velocity of the solution point track is consistent with the start vector If the point track in matches, the judgment flag f=2 will be set to allow the track to start.
Claims
1. A track initiation method based on time-sharing blind spot filling strategy and multi-level information matching, characterized in that: Step 1: Process the data after pulse pressure modulus calculation, calculate the folding times p for each distance unit according to the correlation between short-period pulses, and mark them. The distance unit parameters are statistically calculated in units of pulses as shown in equations (1) and (2); d' i ={δ1,δ2,…,δ m } (1) δ j ={a,u,r δ ,p,t,v,f}; (2) Among them, a, u, r δ , p, t, v and f are the position, distance unit, defuzzified point distance, folding times, time, speed and mark of the jth distance unit that meets the criteria, i = 1, 2, ..., n, n is the number of short pulses in the pulse group, j = 1, 2, ..., m, m is the number of distance units in the pulse that meets the criteria; Step 2: Calculate the Doppler velocity of the corresponding distance unit according to the correlation between short-period pulses, as shown in formula (3): v' di ={v d1 ,v d2 ,…,v dm } (3) And record it in the corresponding distance unit vector δ j In the velocity attribute, j is the index of the distance unit that meets the criteria within the pulse; Step 3: Perform pulse group accumulation, constant false alarm processing and point trace detection condensation. δ and u δ are the direction and distance of the point trace after condensation; Step 4: According to the distance dividing point between the long and short cycles, the point traces of the markers within the short cycle pulse distance segment are defuzzified, as shown in equations (4) and (5): t r =pT+t' (4) R=(c*t r ) / 2 (5) Among them, T is the short wave pulse repetition period, t' is the short pulse delay, and the solved R is stored in the vector δ j r δ ; Step 5: The point traces within the distance segment of the long and short period pulses are recorded as shown in formula (6): c k ={a γ ,r γ ,t γ }; (6) Where k = 1, 2, ..., L, L is the number of points in the long pulse distance segment, in order to confirm the calculated target distance r' δ is the folding target, the direction is a δ The distance is traversed from the boundary point of long and short pulses to the maximum range, and the point traces in the long-period pulse distance segment and the calculated target point traces meet the optimal solution d of the Euclidean distance criterion. min , as shown in formula (7): Among them, d min =min(d), record the vector δ' corresponding to the optimal solution at this time j , and set the discrimination flag f = 1; Step 6: δ' j The point with the discriminant flag f = 1 is used as the candidate point for the start of the track and recorded in the starting point vector, as shown in formula (8): Where l = 1, 2, ..., l max , l max is the number of traces that can be started within the radar cycle, a, d, e, t are the azimuth, distance, elevation and time of the current trace respectively; Step 7: Repeat steps 1 to 6 to calculate the Doppler velocity component v for the solved point traces. dx and v dy , and the displacement component and As shown in equations (9), (10), (11) and (12): in dx =v*cos(a δ ) (9) v dy =v*sin(a δ ) (10) Among them, T r is the radar period; Substitute the displacement component into the formula ρ=xcos(θ)+ysin(θ) to calculate the Doppler velocity displacement corresponding to According to the track start criteria, determine whether the inter-cycle displacement calculated by the Doppler velocity of the solution point track is consistent with the start vector If the point track in matches, the judgment flag f=2 will be set to allow the track to start.
