Track processing method under multiple observation platforms

Through the track processing method of multi-observation platforms, the data processing center is used to filter and correlate the measurement data, which solves the problem of multi-objective track establishment error under a single-observation platform, and achieves higher track accuracy and multi-objective Tracking accuracy.

CN119043337APending Publication Date: 2024-11-29CHANGAN AUTOMOBILE (GRP) CO LTD
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202411443318.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-16
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

Single observation platforms are prone to measurement data errors when dealing with multiple targets, resulting in huge deviations in target track establishment and affecting subsequent work.

Method used

The track processing method of multiple observation platforms is adopted, and the secondary track points are established through the joint fusion of multiple observation platforms, and the data processing center is used to filter and correlate the measurement data, and the normal quality, separation and divorce quality are increased to improve the quality of Accuracy of track establishment.

Benefits of technology

It effectively avoids accuracy errors, improves the accuracy of track establishment and multi-object tracking accuracy, and improves the reliability and real-timeness of track processing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119043337A_ABST
    Figure CN119043337A_ABST
Patent Text Reader

Abstract

A track processing method under multiple observation platforms comprises the following steps: 1) multiple observation platforms respectively acquire measurement data, and each measurement data is transmitted to a data processing center; 2) the data processing center carries out primary track point filtering on all measurement data; 3) if the data processing center does not have any secondary track point, taking the first primary track point entering the data processing center as a secondary track point; if a secondary track point exists in the data processing center, the primary track point of the measurement data of each platform is associated with the secondary track point of the data processing center, and if the primary track point of a certain target is associated with a certain secondary track point, the primary track point and the secondary track point are fused; and if a certain primary track point is not associated with all the secondary track points, adding 1 to the departure quality of the target, and if the departure quality of a certain target is greater than a set departure quality threshold value, establishing the primary track point of the target as a new secondary track point.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of electronic information processing, and in particular to a track processing method under multiple observation platforms. Background Art

[0002] Existing track processing generally uses a single observation platform to collect measurement data. However, due to the increasing number of targets that need to be monitored in the modern environment, the measurement data collected by a single observation platform is getting larger and larger, and the accuracy of a single observation platform is limited. Under certain conditions, the measurement data will have large errors or even lose measurement data, resulting in huge deviations in the establishment of the target track, which seriously affects subsequent work.

[0003] In the invention patent "An Improved Multi-target Radar Track Processing Method" with application publication number CN110031834A, the radar collects target measurement data and tracks it through Kalman filtering, processes the track start, clears invalid tracks, and then establishes a new track. In this method, it is still a single observation platform, that is, a single radar establishes tracks for multiple targets. The more targets there are, the more likely it is that measurement data errors will occur, which will lead to huge deviations in the final target track establishment. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention proposes a track processing method under multiple observation platforms, which avoids the problem of precision error through multi-observation platform processing, and establishes secondary track points by jointly fusing multiple observation platforms, thereby ensuring the accuracy of target track establishment.

[0005] The technical solution of the present invention is: a track processing method under multiple observation platforms, comprising the following steps:

[0006] 1) Multiple observation platforms collect measurement data respectively, and the measurement data collected by each platform is transmitted to the same data processing center. The measurement data includes the collection target, a track point, a timestamp and the station number of the observation platform;

[0007] 2) The data processing center performs filtering on the primary track points of the measurement data of each platform according to the station number of each observation platform;

[0008] 3) Establishment and association of secondary track points;

[0009] If the data processing center does not have any secondary track points, the first primary track point that enters the data processing center will be used as the secondary track point; if the data processing center has secondary track points, the primary track points of the measurement data of each platform will be associated with the secondary track points of the data processing center respectively. If the primary track point of a target is associated with a secondary track point, the primary track point is merged with the secondary track point, and the normalized mass of the target and the secondary track is increased by 1. If the normalized mass of a target and the same secondary track is greater than the set normalized mass threshold, a normalized mapping relationship is established between the target and the secondary track; if a primary track point is not associated with all secondary track points, the divergent mass of the target is increased by 1. If the divergent mass of a target is greater than the set divergent mass threshold, the primary track point of the target is established as a new secondary track point.

[0010] Preferably, in step 2), the data processing center pre-sets corresponding filtering parameters according to the accuracy of each observation platform, and calls the filtering parameters corresponding to the observation platform to filter the primary track points of the measurement data according to the station number of each observation platform in the measurement data.

[0011] Preferably, in step 2), the primary track points of the measurement data of each platform are processed by Kalaman filtering.

[0012] Preferably, in step 3), the step of associating the primary track points of the measurement data of each platform with the secondary track points of the data processing center is:

[0013] ① Each measurement data is compared with the time of the last point of each secondary track point through the timestamp;

[0014] ② If the timestamp of a certain measurement data is less than the time of the last point of a certain secondary track point, the negative time correlation algorithm is used to calculate the correlation parameters of the secondary track point. If the timestamp of a certain measurement data is greater than the time of the last point of a certain secondary track point, the positive time correlation algorithm is used to calculate the correlation parameters of the secondary track point. If the timestamp of a certain measurement data is equal to the time of the last point of a certain secondary track point, the same time correlation algorithm is used to calculate the correlation parameters of the secondary track point.

[0015] ③ If the association parameter of a secondary track point is less than the set threshold value, the secondary track point is associated with the corresponding primary track point of the measurement data; otherwise, the secondary track point is not associated with the corresponding primary track point of the measurement data.

[0016] Preferably, in the step 3), if a primary track point of a target is associated with a secondary track point, the primary track point is merged with the secondary track point, and the normalized mass of the target and the secondary track is increased by 1. If the primary track point of the measurement data collected by the observation platform of the primary track point for the same target and which is continuous with the measurement data of the primary track point is still associated with the secondary track point, their normalized mass is further increased by 1. Otherwise, the normalized mass of the target and the secondary track is set to 0.

[0017] Preferably, in the step 3), if a primary track point is not associated with all secondary track points, the separation quality of the target is increased by 1; if the primary track point of the measurement data collected by the observation platform of the primary track point for the same target and which is continuous with the measurement data of the primary track point is still not associated with all secondary track points, the separation quality of the target is further increased by 1; otherwise, the separation quality of the target is set to 0.

[0018] Preferably, in the step 3), after a normalized mapping relationship is established between a certain target and a certain secondary track, if the primary track point of other measurement data collected by the observation platform of the primary track point for the target is not associated with the secondary track point, then the separation quality of the target and the secondary track point is increased by 1; if the primary track point of the measurement data collected by the observation platform of the primary track point for the same target and which is continuous with the measurement data of the primary track point is still not associated with the secondary track point, then the separation quality of the target is further increased by 1, otherwise the separation quality of the target is set to 0; if the separation quality of the target is greater than the set separation quality threshold, the normalized mapping relationship between the target and the secondary track is deleted.

[0019] Preferably, in step 2), the data processing center unifies the dimensions of the measurement data collected by each observation platform.

[0020] The advantages of the present invention are:

[0021] 1. The present invention sets different filtering parameters according to the accuracy of different observation platforms, which can more accurately eliminate the error deviation points of the primary track points of each measurement data.

[0022] 2. The present invention can handle some special track situations, such as batching and bifurcation, by increasing the normalized mass, separating the mass and the divorced mass. The changes of the three can also provide a basis for whether to associate and merge the tracks, which greatly improves the accuracy of track establishment. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0024] See also Figure 1, a track processing method under multiple observation platforms, comprising the following steps:

[0025] 1) Multiple observation platforms collect measurement data respectively, and the measurement data collected by each platform is transmitted to the same data processing center. The measurement data includes the collection target, a track point, a timestamp and the station number of the observation platform;

[0026] 2) The data processing center performs filtering processing on the primary track points of the measurement data of each platform according to the station number of each observation platform; the data processing center pre-sets the corresponding Kalaman filter parameters according to the accuracy of each observation platform, and the data processing center calls the Kalaman filter parameters corresponding to the observation platform according to the station number of each observation platform in the measurement data to perform Kalaman filter processing on the primary track points of the measurement data. The data processing center also needs to unify the dimensions of the measurement data collected by each observation platform, for example, converting the mils in the angle into radians.

[0027] 3) Establishment and association of secondary track points;

[0028] If the data processing center does not have any secondary track points, the first primary track point that enters the data processing center is used as the secondary track point; if the data processing center has secondary track points, the primary track points of the measurement data of each platform are associated with the secondary track points of the data processing center. The steps of associating the primary track points of the measurement data of each platform with the secondary track points of the data processing center are:

[0029] ① Each measurement data is compared with the time of the last point of each secondary track point through the timestamp;

[0030] ② If the timestamp of a certain measurement data is less than the time of the last point of a certain secondary track point, the negative time correlation algorithm is used to calculate the correlation parameters of the secondary track point. If the timestamp of a certain measurement data is greater than the time of the last point of a certain secondary track point, the positive time correlation algorithm is used to calculate the correlation parameters of the secondary track point. If the timestamp of a certain measurement data is equal to the time of the last point of a certain secondary track point, the same time correlation algorithm is used to calculate the correlation parameters of the secondary track point.

[0031] The same time correlation algorithm is:

[0032]

[0033] In the formula, is the state estimate of a track point of an observation platform at time k, is the covariance matrix corresponding to the state estimate of a track point of an observation platform at time k, is the state estimate of a certain secondary track point, and “[●]” indicates Abbreviation of .

[0034] The positive time correlation algorithm is:

[0035]

[0036] in

[0037]

[0038] In the formula, Φ(k 1 |k 2 ) represents the state transfer matrix, Q(k 2 ) represents the noise matrix, are the state error covariance between the primary and secondary track points respectively.

[0039] The negative time correlation algorithm is:

[0040]

[0041] in

[0042]

[0043] In the formula, Φ(k 2 |k 1 ) represents the state transfer matrix, Q(k 1 ) represents the noise matrix, are the state error covariance between the primary and secondary track points respectively.

[0044] ③ If the association parameter of a secondary track point is less than the set threshold value, in this embodiment, the threshold value is set to 8000, then the secondary track point is associated with the corresponding primary track point of the measurement data, otherwise, the secondary track point is not associated with the corresponding primary track point of the measurement data.

[0045] If a primary track point of a target is associated with a secondary track point, the primary track point is merged with the secondary track point, and the normalized quality of the target and the secondary track is increased by 1. If the primary track point of the measurement data collected by the observation platform of the primary track point for the same target and continuous with the measurement data of the primary track point is still associated with the secondary track point, their normalized quality is increased by 1, otherwise the normalized quality of the target and the secondary track is set to 0. If the normalized quality of a target and the same secondary track is greater than the set normalized quality threshold, in this embodiment, the normalized quality threshold is set to 3, then the target and the secondary track establish a normalized mapping relationship;

[0046] The fusion of the primary track point and the secondary track point is achieved by a simple convex combination fusion method, and the formula is as follows:

[0047]

[0048] P -1 =(P r,i ) -1 +(P j ) -1

[0049] In the formula, Respectively represent the status information of the primary track point and the secondary track point, P r,i , P j Represent the covariance matrices of the primary track point and the secondary track point respectively, and the fusion result is P.

[0050] The formula for the normalized mass is:

[0051]

[0052] In the formula, KR k (r,i,j) is the current normalized mass, KR k-1 (r,i,j) is the previous normalized mass, r is the station number of the observation platform, i is the target, and j is the number of the secondary track point.

[0053] After establishing a normalized mapping relationship, the primary track points of the measurement data collected by the observation platform for the target and continuous with the measurement data of the primary track point are directly associated with the secondary track, and there is no need to calculate the associated parameters. If a normalized mapping relationship is established between a certain target and a certain secondary track, if the primary track points of other measurement data collected by the observation platform of the primary track point for the target are not associated with the secondary track point, the separation quality of the target and the secondary track point is increased by 1, and if the primary track points of the measurement data collected by the observation platform of the primary track point for the same target and continuous with the measurement data of the primary track point are still not associated with the secondary track point, the separation quality of the target is increased by 1, otherwise the separation quality of the target is set to 0; if the separation quality of the target is greater than the set separation quality threshold, and the separation quality threshold is set to 3 in this embodiment, the normalized mapping relationship between the target and the secondary track is deleted; at the same time, the normalized quality or separation quality of the primary track point and other secondary track points is calculated again.

[0054] The calculation formula of the detachment quality is:

[0055]

[0056] KT k (r,i,j) is the current normalized mass, KTk-1 (r,i,j) is the previous normalized mass, r is the station number of the observation platform, i is the target, and j is the number of the secondary track point.

[0057] If a primary track point is not associated with any secondary track points, the separation quality of the target is increased by 1. If the primary track point of the measurement data collected by the observation platform of the primary track point for the same target and which is continuous with the measurement data of the primary track point is still not associated with any secondary track points, the separation quality of the target is increased by 1 again. Otherwise, the separation quality of the target is set to 0. If the separation quality of a target is greater than the set separation quality threshold, in this embodiment, the separation quality threshold is set to 3, and the primary track point of the target is established as a new secondary track point.

[0058] The calculation formula of the detachment quality is:

[0059]

[0060] KN k (r,i,j) is the current normalized mass, KN k-1 (r,i,j) is the previous normalized mass, r is the station number of the observation platform, i is the target, and j is the number of the secondary track point.

[0061] In addition, after each track fusion is completed, the current secondary track point will remove the expired mapping relationship according to the preset time threshold. Similarly, if the secondary track point has not updated data for a long time, that is, there is no new primary track point to fuse with it, it means that the secondary track point may have been disconnected or invalid, and it will also be cleared regularly. These two measures enable the system to obtain the latest fusion information and ensure good real-time performance.

[0062] This method can output the fused track data in the form of real-time transmission or file. Overall, the track processing method under the multi-observation platform of the present invention has a complete track processing flow, which can effectively improve the multi-target tracking accuracy. The algorithm has good reliability and real-time performance, and can fully meet the needs of engineering applications.

Claims

1. A track processing method under multiple observation platforms, characterized in that: The following steps are involved: 1) Multiple observation platforms collect measurement data respectively, and the measurement data collected by each platform is transmitted to the same data processing center. The measurement data includes the collection target, a track point, a timestamp and the station number of the observation platform; 2) The data processing center performs filtering on the primary track points of the measurement data of each platform according to the station number of each observation platform; 3) Establishment and association of secondary track points; If the data processing center does not have any secondary track points, the first primary track point that enters the data processing center will be used as the secondary track point; if the data processing center has secondary track points, the primary track points of the measurement data of each platform will be associated with the secondary track points of the data processing center respectively. If the primary track point of a target is associated with a secondary track point, the primary track point is merged with the secondary track point, and the normalized mass of the target and the secondary track is increased by 1. If the normalized mass of a target and the same secondary track is greater than the set normalized mass threshold, a normalized mapping relationship is established between the target and the secondary track; if a primary track point is not associated with all secondary track points, the divergent mass of the target is increased by 1. If the divergent mass of a target is greater than the set divergent mass threshold, the primary track point of the target is established as a new secondary track point.

2. The track processing method under multiple observation platforms according to claim 1 is characterized in that: In the step 2), the data processing center pre-sets corresponding filtering parameters according to the accuracy of each observation platform. The data processing center calls the filtering parameters corresponding to the observation platform according to the station number of each observation platform in the measurement data to filter the primary track points of the measurement data.

3. The track processing method under multiple observation platforms according to claim 1 is characterized in that: In the step 2), the primary track points of the measurement data of each platform are processed by Kalaman filtering.

4. The track processing method under multiple observation platforms according to claim 1, characterized in that: In step 3), the steps of associating the primary track points of the measurement data of each platform with the secondary track points of the data processing center are: ① Each measurement data is compared with the time of the last point of each secondary track point through the timestamp; ② If the timestamp of a certain measurement data is less than the time of the last point of a certain secondary track point, the negative time correlation algorithm is used to calculate the correlation parameters of the secondary track point. If the timestamp of a certain measurement data is greater than the time of the last point of a certain secondary track point, the positive time correlation algorithm is used to calculate the correlation parameters of the secondary track point. If the timestamp of a certain measurement data is equal to the time of the last point of a certain secondary track point, the same time correlation algorithm is used to calculate the correlation parameters of the secondary track point. ③ If the association parameter of a secondary track point is less than the set threshold value, the secondary track point is associated with the corresponding primary track point of the measurement data; otherwise, the secondary track point is not associated with the corresponding primary track point of the measurement data.

5. The track processing method under multiple observation platforms according to claim 1 is characterized in that: In the step 3), if a primary track point of a target is associated with a secondary track point, the primary track point is merged with the secondary track point, and the normalized mass of the target and the secondary track is increased by 1. If the primary track point of the measurement data collected by the observation platform of the primary track point for the same target and which is continuous with the measurement data of the primary track point is still associated with the secondary track point, their normalized mass is increased by 1 again, otherwise the normalized mass of the target and the secondary track is set to 0.

6. The track processing method under multiple observation platforms according to claim 1 is characterized in that: In the step 3), if a primary track point is not associated with any secondary track points, the separation quality of the target is increased by 1; if the primary track point of the measurement data collected by the observation platform of the primary track point for the same target and which is continuous with the measurement data of the primary track point is still not associated with any secondary track points, the separation quality of the target is further increased by 1; otherwise, the separation quality of the target is set to 0.

7. The track processing method under multiple observation platforms according to claim 1 is characterized in that: In the step 3), after a normalized mapping relationship is established between a certain target and a certain secondary track, if the primary track point of other measurement data collected by the observation platform of the primary track point for the target is not associated with the secondary track point, then the separation quality of the target and the secondary track point is increased by 1; if the primary track point of the measurement data collected by the observation platform of the primary track point for the same target and which is continuous with the measurement data of the primary track point is still not associated with the secondary track point, then the separation quality of the target is increased by 1 again, otherwise the separation quality of the target is set to 0; if the separation quality of the target is greater than the set separation quality threshold, the normalized mapping relationship between the target and the secondary track is deleted.

8. The track processing method under multiple observation platforms according to claim 1 is characterized in that: In the step 2), the data processing center unifies the dimensions of the measurement data collected by each observation platform.

Citation Information

Patent Citations

  • Multi-sensor track association method

    CN111767639A

  • Multi-target tracking algorithm based on track management method

    CN112946624A

  • Obstacle prediction and tracking method based on ICP and multi-feature data association

    CN114371485A

  • VTS system-oriented multi-sensor information fusion method

    CN115718905A

  • Target tracking and anti-interference method based on active radar and infrared sensor

    CN116973908A