A method for correlating active sonar multi-target point-tracks

Abstract

The present invention relates to the technical field of underwater acoustic signal processing, and in particular to an active sonar multi-target point track-track association method. It includes the following steps: Step 1, track establishment: all point tracks in the initial period are assigned new track numbers. The measured point tracks that cannot be associated with the established tracks are assigned new tracks M+1 to form new track numbers. Step 2, point track-track association: the track already has less than 3 points: determine whether the change in the position information of the detected point track and the latest historical point track on the track is within a threshold. If it is within the threshold, give the current detected point track and track a membership evaluation value; Step 3, track temporary storage; Step 4, track cancellation. The present invention solves the problems of complex tracking systems and large amount of calculations, is easier to implement in engineering, and has strong versatility. For different types of sonars, it is only necessary to improve the corresponding thresholds according to system errors and screening target types.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater acoustic signal processing, and is mainly used for multi-target automatic tracking and track management technology in active sonar applications, and aims to obtain target real motion information through point track-track association to provide a basic basis for target identification, and particularly relates to an active sonar multi-target point track-track association method. Background Art

[0002] The main function of point-track association is to complete the update and maintenance of the track, that is, to determine the predicted value of the track as the center, select the measured points that meet the conditions according to certain specific criteria, and use the stored points in the track for filtering.

[0003] Multi-target point-track-track association technology has always been a research hotspot. When the target is affected by natural and human factors such as strong clutter and interference, it is difficult to effectively distinguish real data from contaminated data, especially when the real data of the target is lost, the difficulty of association is greatly increased. Without considering traditional data association problems such as false alarms and missed alarms, more complex problems such as data confusion and crossover may occur due to high maneuvers and interactive movements of the target. This is also an urgent problem to be solved in the field of multi-target point-track-track association. The effective management of point-track-track association and target tracks is more conducive to the accurate acquisition of target motion characteristics and provides a solid foundation for target identification.

[0004] However, with the increasing demand for intelligent sonar equipment, automatic target tracking technology also needs to be improved. In the actual use of sonar, there may be complex situations such as target merging and crossing within the detection area, and due to errors in direction finding and distance measurement, the changes in target detection points cannot strictly conform to the motion model. In the above complex situations, the track cannot be maintained and is greatly affected by clutter interference, resulting in confusion in track management results. Summary of the invention

[0005] The purpose of the present invention is to provide an active sonar multi-target point track-track association method. The present invention utilizes the comprehensive utilization of historical tracks and Kalman prediction results, obtains a similarity matrix through the nearest neighbor principle, and improves the situation where the track cannot be maintained in complex situations such as target merging and crossing.

[0006] In order to solve the above technical problems, the present invention provides an active sonar multi-target point track-track association method, comprising the following steps:

[0007] Step 1: Track establishment:

[0008] All measured points in the initial period are assigned new track numbers The measured point track that cannot be associated with the established track is assigned a new track M+1 to form a new track number

[0009] Step 2: Point-to-track association:

[0010] If the track has less than or equal to 2 points, determine whether the position information change between the measured point and the latest historical point on the track is within the threshold. If it is within the threshold, give the current measured point and track a membership evaluation value; If the track has more than 2 points, use the Kalman filter to predict the latest position information of the track, determine whether the position information change between the measured point and the predicted track value is within the threshold. If it is within the threshold, give the current measured point and track a membership evaluation value;

[0011] Step 3: Temporary storage of tracks:

[0012] The track associated with the measured point track of the current period is saved; when the track is not obtained, the measured point track is updated, and the track point is supplemented, that is, the Kalman prediction value of the last point track of the track is used as the updated point track, the track is saved, and return to step 1;

[0013] Step 4: Cancel the track:

[0014] When the number of point updates reaches the track extinction threshold, the track is extinct and all information about the track is deleted.

[0015] Preferably, in step 2, the following conditions must also be met:

[0016] Nearest neighbor principle: the nearest neighbors between the target and the existing trajectory are preferably paired;

[0017] Neighborhood threshold: Whether the target and the track can match must meet the threshold condition based on physical reality. The threshold is the maximum azimuth change rate of the target Δθ max and the distance change rate ΔR max =V max T; where V max is the maximum speed of the target of interest, T is the processing time of the current cycle;

[0018] Mapping principle: A target point can and must match an existing or newly created trajectory, and a trajectory can only correspond to a newly detected target position or there is no target corresponding to it.

[0019] Preferably, the membership evaluation value of any measured point track and any track in step 2 is expressed as:

[0020]

[0021] Where: α, β are weighting coefficients, ΔR is the distance change rate of the point trace, ΔR max is the distance change rate threshold, Δθ is the azimuth change rate of the point trace, Δθmax is the azimuth change rate threshold.

[0022] Preferably, the track number corresponding to the minimum membership value is the track to which the measured point belongs.

[0023] Preferably, the membership evaluation value ranges from [0-1], and the smaller the value, the higher the membership.

[0024] Preferably, the track extinction threshold condition in step 4 is: if the track fails to find the belonging measured point for three consecutive cycles, the track is considered to be extinct.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] 1) The present invention utilizes the historical track and the Kalman prediction results in a comprehensive manner, and obtains the similarity matrix through the nearest neighbor principle to improve the situation where the track cannot be maintained in complex situations such as target merging and crossing.

[0027] 2) The present invention solves the problems of complex tracking systems and large amount of calculations, is easier to implement in engineering, and has strong versatility. For different types of sonars, it is only necessary to improve the corresponding threshold according to the system error and the type of target to be screened. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a processing flow chart of the method provided by an embodiment of the present invention.

[0029] Figure 2 It is a schematic diagram of theoretical trajectories of targets and clutter provided by an embodiment of the present invention.

[0030] Figure 3 This is a graph of the algorithm track management results provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The advantages and features of the present invention will become more apparent from the following description. It should be noted that the accompanying drawings are in very simplified form and in non-precise proportions, and are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention.

[0032] like Figure 1 As shown, an embodiment of the present invention provides an active sonar multi-target point track-track association method, which specifically includes the following:

[0033] Step S1: If it is the start cycle, assign new track numbers to all measured points. Track number The following is expressed: Track 1: {(θ 11 ,R 11 ),...(θ1n , R 1n )} n < N,..... Track M: {(θ M1 , R M1 ),...(θ Mn , R Mn )} n < N; If it is the Nth cycle (N > 1), then the track M+1 generated by returning step S4: {(θ M1 , R M1 ),...(θ Mn , R Mn )} n < N is added to the original track number to form a new track number

[0034] When entering the starting cycle and the Nth cycle as described above, the number of track points included in each existing track is judged. If the number of track points in each track is less than or equal to 2, then the azimuth and distance differences between the measured point (θ, R) and the latest track point (θ Mn , R Mn )(n < N) are directly calculated; if the number of track points is greater than 2, then the azimuth and distance differences between the measured point (θ, R) and the track point predicted by kalman filtering are calculated;

[0035] Step S2, judge whether the azimuth and distance differences between the measured point and the track point / predicted track point are greater than the preset threshold. If they are greater than the preset threshold, then repeatedly judge the azimuth and distance differences between the measured point (θ, R) and the historical track point of the track (θ Mn-1 , R Mn-1 ); otherwise, calculate the membership value of the measured point and the track In the formula: α, β are weighting coefficients, ΔR is the distance change rate of the point trace, ΔR max is the distance change rate threshold, Δθ is the azimuth change rate of the point trace, Δθ max is the azimuth change rate threshold, and it is listed in the membership matrix where n is the number of cycles and m is the track number;

[0036] Step S3, judge the track number corresponding to the minimum membership value. The track corresponding to this track number is the track to which the measured point belongs (that is, the track associated with the measured point trace in the current cycle is saved);

[0037] Step S4, if no track to which the measured point belongs is found, then a new track is assigned to the measured point, track M+1: {(θ M1 , R M1 ),...(θ Mn , R Mn )} n < N, that is, return to step S1 for the next cycle loop;

[0038] Step S5: If the track fails to find the assigned measured point for three consecutive cycles, the track is considered dead and cancelled.

[0039] Simulation conditions: Two tracks are set, one of which is a straight line moving at a speed of 10 kn, and the other is a circular arc moving at a speed of 20 kn. There are scattered clutter interference points around the track.

[0040] like Figure 2 As shown in the figure, the blue line (i.e., track 2 dotted line) is the target track simulating linear motion, the black line (i.e., track 1 solid line) is the target track simulating circular motion, and the red mark (i.e., □) indicates that the target point track in the current cycle has not been detected, and there is a phenomenon of missed detection. Due to the setting of direction finding and range finding errors, the changes in the simulated target detection points cannot strictly conform to the motion model. The green point track (i.e., ○) is a randomly scattered clutter point.

[0041] 1000 Monte Carlo tests were conducted to verify the accuracy of track management results under different complex conditions such as target merging, intersection, direction finding, distance finding, missed detection, etc. The comparison results are shown in Table 1:

[0042] Two tracks are set, each track contains a maximum of 20 points. If the track management result and the track point coincide with each other by more than 50%, the track management is considered correct.

[0043] With the increase of direction finding and ranging errors, the accuracy of track management gradually decreases. For actual sonar equipment (with moderate and below direction finding and ranging errors), the track management results of the algorithm of the present invention are in good agreement with the theoretical track.

[0044] The above description is only a description of the preferred embodiments of the present invention, and is not intended to limit the scope of the present invention. Any changes or modifications made by a person skilled in the art in the field of the present invention based on the above disclosure shall fall within the scope of protection of the claims.

Claims

1. An active sonar multi-target point track-track association method, characterized in that: The steps include: Step 1: Track establishment: All measured points in the initial period are assigned new track numbers ; The measured point track that cannot be associated with the established track is assigned a new track M+1 to form a new track number ; Step 2: Point-to-track association: If the track has less than or equal to 2 points, determine whether the position information change between the measured point and the latest historical point on the track is within the threshold. If it is within the threshold, give the current measured point and track a membership evaluation value; If the track has more than 2 points, use the Kalman filter to predict the latest position information of the track, determine whether the position information change between the measured point and the predicted track value is within the threshold. If it is within the threshold, give the current measured point and track a membership evaluation value; Step 3: Temporary storage of tracks: The track associated with the measured point track of the current period is saved; when the track is not obtained, the measured point track is updated, and the track point is supplemented, that is, the Kalman prediction value of the last point track of the track is used as the updated point track, the track is saved, and return to step 1; Step 4: Cancel the track: When the number of point-filling updates reaches the track extinction threshold, the track is extinct and all information about the track is deleted. In step 2, the following conditions must also be met: Nearest neighbor principle: the two nearest neighbor points between the target point and the existing trajectory are optimally paired; Neighborhood threshold: Whether the target and the track can match must meet the threshold conditions based on physical reality. The threshold is the maximum azimuth change rate of the target. and the rate of change of distance ; In the formula, The maximum speed of the target of interest, Processing time for the current cycle; Mapping principle: A target point can and must match an existing or newly created trajectory, and a trajectory can only correspond to a newly detected target position or there is no target corresponding to it.

2. The active sonar multi-target point track-track association method according to claim 1, characterized in that: The membership evaluation value of any measured point track and any track in step 2 is expressed as: ； Where: is the weighting coefficient, is the distance change rate of the point trace, is the distance change rate threshold, is the azimuth change rate of the point trace, is the azimuth change rate threshold.

3. The active sonar multi-target point track-track association method as claimed in claim 2, characterized in that: The membership evaluation value ranges from [0-1], and the smaller the value, the higher the membership.

4. The active sonar multi-target point track-track association method as claimed in claim 3, characterized in that: The track number corresponding to the minimum membership evaluation value is the track to which the measured point belongs.

5. The active sonar multi-target point track-track association method as claimed in claim 1, characterized in that: The track extinction threshold condition in step 4 is: if the track fails to find the assigned measured point for three consecutive cycles, the track is considered to be extinct.

Images