Formation information-based swarm unmanned aerial vehicle radar tracking data interconnection method

Abstract

The invention discloses a swarm unmanned aerial vehicle radar tracking data interconnection method based on formation information, and relates to the technical field of unmanned aerial vehicle radar tracking. The method comprises the following steps: firstly, performing formation detection, namely determining a distance error threshold, calculating a target relative position, an azimuth angle and a pitch angle matrix, judging whether two targets belong to the same formation or not, and repeating the operation to determine the formation conditions of all the targets; then, carrying out auxiliary interconnection, namely determining three relative position matrixes of the target at the data interconnection moment, subtracting the three relative position matrixes from a reference matrix, calculating the number of two elements in a matrix difference matrix of the relative pitch angle and the azimuth angle of the target, and summing the relative position matrixes of the target corresponding to the minimum matrix; and taking a matrix corresponding mode with the minimum sum as a final data interconnection mode, then updating the reference matrix, and deleting the terminated track. The method effectively solves the problem that radar measurement data is easy to be mislinked due to small size and short distance of the bee colony unmanned aerial vehicle, and improves the tracking accuracy.

Description

Technical Field

[0001] The invention relates to the technical field of unmanned aerial vehicle radar tracking, and in particular to a swarm of unmanned aerial vehicle radar tracking data interconnection method based on formation information. Background Art

[0002] The vigorous development of information technology and industrial manufacturing technology has opened a new era of rapid development for drones. In the military field, drones have performed particularly well, often working together in groups and each taking on different tasks. This tactic is called swarm drone tactics. Compared with manned aircraft, swarm drones have significant advantages such as low cost and excellent penetration capabilities, and have been widely used in modern warfare. Therefore, in the face of enemy drone threats, one must not only actively use swarm drones to launch attacks, but also strengthen defense against them.

[0003] From a technical perspective, radar tracking of swarm drones is essentially a radar multi-target tracking problem. During the multi-target tracking process, the radar will obtain measurement data of multiple targets at each moment. Only by accurately matching the measurement data of the same target at two adjacent moments can the subsequent state estimation steps be carried out smoothly. It can be said that data interconnection is the most critical link in the entire radar multi-target tracking process. Once an error occurs in data interconnection, even if the subsequent state estimation algorithm performs well, the final track result will be very different from the actual situation.

[0004] At present, the data interconnection algorithms for multi-target tracking are mainly divided into two categories. The first category is the maximum likelihood data interconnection algorithm, which is based on the likelihood ratio of the observed data. In actual operation, it needs to calculate the likelihood function of each possible track, and delete the tracks whose likelihood function is lower than a certain threshold, and only retain the rest. This process is computationally intensive, and since it was proposed in the mid-twentieth century, it has been lacking major improvements, so it is rarely used in practical applications; the second category is the Bayesian data interconnection algorithm, which uses the Bayesian criterion as the cornerstone for data interconnection, has a small amount of calculation, and is easy to apply in engineering practice. Common Bayesian data interconnection algorithms include the nearest neighbor data interconnection algorithm (NNSF), the global nearest neighbor data interconnection algorithm (KNNSF), and the joint probability data interconnection algorithm (JPDA). However, when tracking swarm drones, due to the small size of drones, the distance between multiple drones when flying is close to each other, and the distance between the measurement data obtained by the radar is also close. This special situation makes the existing data interconnection algorithm very easy to misconnect. Summary of the invention

[0005] The purpose of the present invention is to propose a swarm UAV radar tracking data interconnection method based on formation information, which is divided into two stages: formation detection and auxiliary interconnection. It effectively solves the problem that radar measurement data is easily misconnected due to the small size and close spacing of swarm UAVs, and improves tracking accuracy.

[0006] To achieve the above purpose, the present invention proposes a swarm UAV radar tracking data interconnection method based on formation information, and the specific steps are as follows:

[0007] Step S1, formation detection, judging the target formation relationship through the relative position, azimuth and pitch angle matrix difference at consecutive moments;

[0008] Step S2: auxiliary interconnection, combining the difference between the candidate interconnection matrix and the reference matrix, and dynamically optimizing the data association result.

[0009] Preferably, in step S1, the formation detection step is as follows:

[0010] Step S11, determining a distance error threshold s, the distance error threshold s is related to the distance of the target radar, the longer the distance, the greater the value of the distance error threshold s;

[0011] Step S12, calculating the target relative position matrix, target relative azimuth matrix and target relative pitch angle matrix at three consecutive moments;

[0012] Step S13, determining whether two targets belong to the same formation by using the three target relative matrices at three consecutive moments calculated in step S12;

[0013] Step S14: Repeat step S13 to obtain the formation status of all targets.

[0014] Preferably, in step S12, D t (i,j),Am t (i,j) and Pi t (i, j) is the relative position relationship between target i and target j at time t. Assume that the coordinates of target i and target j after the state estimation stage at time t are and Then the relative position matrix D of target i and target j is t The calculation formula for (i,j) is:

[0015]

[0016] Where t is time, r i is the radius of target i, r j is the radius of target j; is the azimuth of target i, is the azimuth of target j; θ tis the pitch angle of target i, θ j is the pitch angle of target j;

[0017] Azimuth The range of is 0°~360°, and it is divided into 360 zones with 1° as the boundary, and the zone numbers start from 0 to 359. Then the relative azimuth matrix Am of target i and target j is t The calculation formula for (i,j) is:

[0018]

[0019] Among them, amr i for Area code, amr j for Area code;

[0020] The pitch angle θ ranges from 0° to 90° and is divided into 45 zones with a 2° interval, with zone numbers starting from 0 to 44. The relative pitch angle matrix Pi of target i and target j is t The calculation formula for (i,j) is:

[0021]

[0022] Among them, pri i is θ i Area code, pri j is θ j The area code where you are located.

[0023] Preferably, in step S13, D t-2 , D t-1 , D t Subtract the two and take the absolute value and Am t-2 、Am t-1 、Am t Subtract the two and take the absolute value and Pi t-2 、Pi t-1 、Pi t Subtract the two and take the absolute value and If the relationship between target i and target j is satisfied

[0024] If these 9 conditions are met, then target i and target j are confirmed to be in the same formation, and target i and target j are added to the same formation record;

[0025] Among them, Dt-2 is the target relative position matrix at time t-2, D t-1 is the target relative position matrix at time t-1, D t is the target relative position matrix at time t, D t-2 and D t-1 The absolute value of the difference, D t-1 and D t The absolute value of the difference, D t-2 and D t The absolute value of the difference; Am t-2 is the target relative azimuth matrix at time t-2, Am t-1 is the target relative azimuth matrix at time t-1, Am t is the target relative azimuth matrix at time t, for Am t-2 and Am t-1 The absolute value of the difference, for Am t-1 and Am t The absolute value of the difference, for Am t-2 and Am t The absolute value of the difference; Pi t-2 is the target relative pitch angle matrix at time t-2, Pi t-1 is the target relative pitch angle matrix at time t-1, Pi t is the target relative pitch angle matrix at time t, For Pi t-2 and Pi t-1 The absolute value of the difference, For Pi t-1 and Pi t The absolute value of the difference, For Pi t-2 and Pi t The absolute value of the difference.

[0026] Preferably, in step S2, the auxiliary interconnection step is as follows:

[0027] Step S21, calculating three relative position matrices of all targets at the time of data interconnection;

[0028] Step S22, subtracting the three relative position matrices obtained in step S21 from the three reference relative position matrices;

[0029] Step S23, calculating the number of 2 elements in the difference matrix of the target relative pitch angle and relative azimuth angle matrix, if there is only one matrix with the smallest number, directly taking the data interconnection mode corresponding to the matrix as the final data interconnection mode, skipping step S24;

[0030] Step S24, calculating the sum of the target relative position matrices corresponding to the matrix with the smallest number of 2 elements in the matrix difference matrix of all target relative pitch angles and relative azimuth angles, and taking the data interconnection mode corresponding to the matrix with the smallest sum as the final data interconnection mode;

[0031] Step S25, updating the reference target relative position matrix;

[0032] Step S26, deleting the terminated tracks in the formation record matrix.

[0033] Preferably, in step S21, it is assumed that at time t there are Q candidate data interconnection situation matrices e1, e2…e Q , respectively calculate the target relative position matrix at that moment after state estimation using each data interconnection matrix as the final data interconnection mode Target relative azimuth matrix and the target relative pitch angle matrix in, The target relative position matrix, target relative azimuth matrix, and target relative elevation matrix at the moment after state estimation is performed using the qth candidate data interconnection situation matrix as the final data interconnection mode;

[0034] Preferably, in step S22, each target relative position matrix Target relative azimuth matrix and the target relative pitch angle matrix Relative position matrix D of the reference target b , the relative azimuth matrix of the reference target Am b and the reference target relative pitch angle matrix Pi b Make a difference and take the absolute value and in, for With D b The absolute value of the difference, for With Am b The absolute value of the difference, for With Pi b The absolute value of the difference.

[0035] Preferably, in step S24, each and The number of elements whose value is 2 is denoted by num q , if all num q There are multiple minimum values ​​num u ,num u+1 ,…num o , then the corresponding Sum all elements of the matrix separately to get From e u ,e u+1 …e o Winning Order The smallest e o As the final data interconnection method; among them, e u ,e u+1 ,e o They are the interconnection matrices of the uth, u+1th, and oth candidate data respectively.

[0036] Preferably, in step S25, after determining the final data interconnection mode, it is necessary to calculate the relative position matrix D of the reference target. b , the relative azimuth matrix of the reference target Am b and the reference target relative pitch angle matrix Pi b Update the updated reference target relative position matrix D b ′、Relative azimuth matrix of reference target Am b ′ and the reference target relative pitch angle matrix Pi b ′ are respectively:

[0037]

[0038] in, and The target relative position matrix, target relative azimuth matrix, and target relative elevation matrix are respectively the target relative position matrix, target relative azimuth matrix, and target relative elevation matrix at the moment after the state estimation is performed using the Oth candidate data interconnection situation matrix as the final data interconnection mode.

[0039] Therefore, the present invention proposes a swarm UAV radar tracking data interconnection method based on formation information, and its beneficial effects are as follows:

[0040] (1) The method for interconnecting radar tracking data of swarm drones based on formation information proposed in the present invention effectively reduces the probability of misconnection by combining formation information for data interconnection, improves the accuracy of radar tracking data interconnection, makes the track results more consistent with the actual situation, and solves the problem that radar measurement data is prone to misconnection due to the small size and close spacing of swarm drones;

[0041] (2) In the formation detection phase, the proposed method for interconnecting data of swarm drone radar tracking based on formation information avoids repeated calculations of detected targets by setting rules, thereby reducing the overall calculation amount of the algorithm; in the auxiliary interconnection phase, the final data interconnection method is determined based on specific calculation and comparison methods, thereby improving data processing efficiency;

[0042] (3) The present invention proposes a method for interconnecting swarm UAV radar tracking data based on formation information, which performs well in tracking accuracy, precision and processing efficiency, and improves the tracking accuracy of the radar system for swarm UAV targets.

[0043] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a flow chart of a formation detection algorithm processing of a swarm UAV radar tracking data interconnection method based on formation information of the present invention;

[0045] Figure 2 A formation-assisted data interconnection algorithm flow chart of a swarm UAV radar tracking data interconnection method based on formation information of the present invention;

[0046] Figure 3 is the actual motion trajectory of the target in the embodiment;

[0047] Figure 4 This is the main view of the experimental tracking result in the embodiment;

[0048] Figure 5 It is a top view of the experimental tracking results in the embodiment. DETAILED DESCRIPTION

[0049] In order to make the technical solutions, advantages and purposes of the present invention clearer, the technical solutions of the embodiments of the present invention are clearly and completely described below. The described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the described embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the protection scope of this application.

[0050] Unless otherwise defined, technical or scientific terms used in the present invention shall have the common meanings understood by one having ordinary skills in the field to which the present invention belongs.

[0051] Example

[0052] The target motion trajectory of the tracking target in this embodiment is obtained by simulating the actual flight data of multiple UAVs. Figure 3As shown in the figure, there are six drones in total. The general shape of the flight trajectory of each drone is a serpentine trajectory in three-dimensional space. From the figure, we can see that each drone moves within a range of about 6 to 9 km from the radar, with a flight speed of about 20 m / s. The distance between two adjacent drones is about 50 m. Each drone has been climbing before the 13th radar sampling cycle, and began to gradually descend after the 25th radar sampling cycle. There are obvious turning maneuvers at multiple radar sampling moments. The specific parameters are shown in Table 1.

[0053] Table 1 Specific parameters used in the experiment

[0054]

[0055] The present invention provides a swarm UAV radar tracking data interconnection method based on formation information, which is divided into two stages. The first stage is the formation detection stage, such as Figure 1 As shown, the specific steps are as follows:

[0056] Step 1: Determine the distance error threshold as 50.

[0057] Step 2: Calculate the target relative position matrix D for three consecutive moments after the track starts t , target relative azimuth matrix Am t and the target relative pitch angle matrix Pi t , assuming that the radar is tracking n targets at time t, then these matrices are all matrices with n rows and n columns, D t (i,j),Am t (i,j) and Pi t (i, j) together represent the relative position relationship between target i and target j at time t. They are all upper triangular matrices, that is, only elements at positions j>i are used. Assume that at time t, after the state estimation stage, the coordinates of target i and target j are and Then D t The calculation formula for (i,j) is:

[0058]

[0059] In calculating Am t (i,j) and Pi t (i,j), the radar measurement space needs to be partitioned first. For example, its range is 0°~360°, and it is divided into 360 zones with 1° as the boundary. The zone numbers start from 0 and go to 359. Assume The area code is amr i , The area code is amr j; For the pitch angle θ, its range is 0°~90°, and it is divided into 45 zones with 2° as the boundary, and the zone number starts from 0 to 44. Assume θ i The area code is pri i ,θ j The area code is pir j , then Am t (i,j) and Pi t The calculation formula for (i,j) is:

[0060]

[0061] Step 3: D t-2 , D t-1 , D t Subtract the two and take the absolute value and Am t-2 、Am t-1 、Am t Subtract the two and take the absolute value and Pi t-2 、Pi t-1 、Pi t Subtract the two and take the absolute value and If the i-th goal and the j-th goal satisfy These 9 conditions assume that the ith target and the jth target are in the same formation and add these two targets to the same formation record.

[0062] Step 4: Repeat step 2 until all targets are subjected to a formation detection algorithm, and finally a formation record matrix with 5 rows and 5 columns is obtained.

[0063] The second stage: auxiliary interconnection stage, such as Figure 2 As shown, the specific steps are as follows:

[0064] Step 1: Assume that at time t there are Q candidate data interconnection matrices e1, e2…e Q , respectively calculate the target relative position matrix at that moment after state estimation using each data interconnection matrix as the final data interconnection mode Target relative azimuth matrix and the target relative pitch angle matrix

[0065] Step 2: Relative position matrix of each target Target relative azimuth matrix and the target relative pitch angle matrix Relative position matrix D with reference target b , the relative azimuth matrix of the reference target Am b and the reference target relative pitch angle matrix Pi b Make a difference and take the absolute value and

[0066] Step 3: Calculate each and The number of elements whose value is 2 is denoted by num q , if all num q There is only one minimum value num o , then determine e o For the final data interconnection method, skip step 4 and proceed to step 5.

[0067] Step 4: If all num q There are multiple minimum values ​​num u ,num u+1 ,…num o , then the corresponding Sum all elements of the matrix separately to get From e u ,e u+1 …e o Winning Order The smallest e o As the ultimate data interconnection method.

[0068] Step 5: After determining the final data interconnection method, it is necessary to calculate the relative position matrix D of the reference target. b , the relative azimuth matrix of the reference target Am b and the reference target relative pitch angle matrix Pi b Update the updated reference target relative position matrix D b ′、Relative azimuth matrix of reference target Am b ′ and the reference target relative pitch angle matrix Pi b ′ are respectively:

[0069]

[0070] Step 6: If some tracks have track termination, if these tracks are in a certain formation record, these corresponding tracks need to be deleted from the formation record.

[0071] like Figure 4-5As shown in Table 2, after the tracking is completed, the root mean square error (RMSE) of the algorithm is 36.6877 when tracking six drones, the multi-target tracking accuracy (MOTA) is 0.9073, and the number of redundant targets (N redundancy ) is 0, and the average processing time is 16.0601 seconds. It performs well in tracking precision, accuracy and processing efficiency, and can better meet the needs of practical applications.

[0072] Table 2 Formation detection and auxiliary interconnection algorithm tracking performance indicators

[0073] RMSE MOTA <![CDATA[N redundancy ]]> t Formation detection and auxiliary interconnection algorithm 36.6877 0.9073 0 16.0601s

[0074] Therefore, the present invention provides a swarm UAV radar tracking data interconnection method based on formation information. By combining formation information for data interconnection, the problem of easy misconnection of radar measurement data due to the small size and close spacing of swarm UAVs is effectively solved, thereby improving the accuracy of radar tracking data interconnection.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solution of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solution to deviate from the spirit and scope of the technical solution of the present invention.

Claims

1. A method for interconnecting radar tracking data of swarm drones based on formation information, characterized in that: The specific steps are as follows: Step S1, formation detection, judging the target formation relationship through the relative position, azimuth and pitch angle matrix difference at consecutive moments; Step S2: auxiliary interconnection, combining the difference between the candidate interconnection matrix and the reference matrix, and dynamically optimizing the data association result.

2. According to the method for interconnecting swarm drone radar tracking data based on formation information of claim 1, it is characterized in that: In step S1, the formation detection steps are as follows: Step S11, determining a distance error threshold s, the distance error threshold s is related to the distance of the target radar, the longer the distance, the greater the value of the distance error threshold s; Step S12, calculating the target relative position matrix, target relative azimuth matrix and target relative pitch angle matrix at three consecutive moments; Step S13, determining whether two targets belong to the same formation by using the three target relative matrices at three consecutive moments calculated in step S12; Step S14: Repeat step S13 to obtain the formation status of all targets.

3. The method for interconnecting radar tracking data of swarm drones based on formation information according to claim 2 is characterized in that: In step S12, D t (i,j),Am t (i,j) and Pi t (i, j) is the relative position relationship between target i and target j at time t. It is assumed that the coordinates of target i and target j after the state estimation stage at time t are and Then the relative position matrix D of target i and target j is t The calculation formula for (i,j) is: Where t is time, r i is the radius of target i, r j is the radius of target j; is the azimuth of target i, is the azimuth of target j; θ i is the pitch angle of target i, θ j is the pitch angle of target j; Azimuth The range of is 0°~360°, and it is divided into 360 zones with 1° as the boundary, and the zone numbers start from 0 to 359. Then the relative azimuth matrix Am of target i and target j is t The calculation formula for (i,j) is: Among them, amr i for Area code, amr j for Area code; The pitch angle θ ranges from 0° to 90° and is divided into 45 zones with a 2° interval, with zone numbers starting from 0 to 44. The relative pitch angle matrix Pi of target i and target j is t The calculation formula for (i,j) is: Among them, pri i is θ i Area code, pir j is θ j The area code where you are located.

4. The method for interconnecting swarm drone radar tracking data based on formation information according to claim 3 is characterized in that: In step S13, D t-2 , D t-1 , D t Subtract the two and take the absolute value and Am t-2 、Am t-1 、Am t Subtract the two and take the absolute value and Pi t-2 、Pi t-1 、Pi t Subtract the two and take the absolute value and If the relationship between target i and target j is satisfied If these 9 conditions are met, then target i and target j are confirmed to be in the same formation, and target i and target j are added to the same formation record; Among them, D t-2 is the target relative position matrix at time t-2, D t-1 is the target relative position matrix at time t-1, D t is the target relative position matrix at time t, D t-2 and D t-1 The absolute value of the difference, D t-1 and D t The absolute value of the difference, D t-2 and D t The absolute value of the difference; Am t-2 is the target relative azimuth matrix at time t-2, Am t-1 is the target relative azimuth matrix at time t-1, Am t is the target relative azimuth matrix at time t, for Am t-2 and AM t-1 The absolute value of the difference, For AM t-1 and Am t The absolute value of the difference, for Am t-2 and Am t The absolute value of the difference; Pi t-2 is the target relative pitch angle matrix at time t-2, Pi t-1 is the target relative pitch angle matrix at time t-1, Pi t is the target relative pitch angle matrix at time t, For Pi t-2 and Pi t-1 The absolute value of the difference, For Pi t-1 and Pi t The absolute value of the difference, For Pi t-2 and Pi t The absolute value of the difference.

5. The method for interconnecting swarm drone radar tracking data based on formation information according to claim 1 is characterized in that: In step S2, the auxiliary interconnection steps are as follows: Step S21, calculating three relative position matrices of all targets at the time of data interconnection; Step S22, subtracting the three relative position matrices obtained in step S21 from the three reference relative position matrices; Step S23, calculating the number of 2 elements in the difference matrix of the target relative pitch angle and relative azimuth angle matrix, if there is only one matrix with the smallest number, directly taking the data interconnection mode corresponding to the matrix as the final data interconnection mode, skipping step S24; Step S24, calculating the sum of the target relative position matrices corresponding to the matrix with the smallest number of 2 elements in the matrix difference matrix of all target relative pitch angles and relative azimuth angles, and taking the data interconnection mode corresponding to the matrix with the smallest sum as the final data interconnection mode; Step S25, updating the reference target relative position matrix; Step S26, deleting the terminated tracks in the formation record matrix.

6. The method for interconnecting swarm drone radar tracking data based on formation information according to claim 5 is characterized in that: In step S21, it is assumed that at time t there are Q candidate data interconnection matrixes e1, e2…e Q , respectively calculate the target relative position matrix at that moment after state estimation using each data interconnection matrix as the final data interconnection mode Target relative azimuth matrix and the target relative pitch angle matrix in, The target relative position matrix, target relative azimuth matrix, and target relative pitch angle matrix are obtained at the moment when state estimation is performed using the qth candidate data interconnection situation matrix as the final data interconnection mode.

7. The method for interconnecting swarm drone radar tracking data based on formation information according to claim 6 is characterized in that: In step S22, each target relative position matrix Target relative azimuth matrix and the target relative pitch angle matrix Relative position matrix D of the reference target b , the relative azimuth matrix of the reference target Am b and the reference target relative pitch angle matrix Pi b Make a difference and take the absolute value and in, for With D b The absolute value of the difference, for With Am b The absolute value of the difference, for With Pi b The absolute value of the difference.

8. The method for interconnecting swarm drone radar tracking data based on formation information according to claim 7 is characterized in that: In step S24, each and The number of elements whose value is 2 is denoted by num q , if all num q There are multiple minimum values ​​num u ,num u+1 ,…num o , then the corresponding Sum all elements of the matrix separately to get From e u ,e u+1 …e o Winning Order The smallest e o As the final data interconnection method; among them, e u ,e u+1 ,e b They are the interconnection matrices of the uth, u+1th, and oth candidate data respectively.

9. The method for interconnecting swarm drone radar tracking data based on formation information according to claim 8 is characterized in that: In step S25, after the final data interconnection method is determined, the reference target relative position matrix D b , the relative azimuth matrix of the reference target Am b and the reference target relative pitch angle matrix Pi b Update the updated reference target relative position matrix D b ′、Relative azimuth matrix of reference target Am b ′ and the reference target relative pitch angle matrix Pi b ′ are respectively: in, and The target relative position matrix, target relative azimuth matrix, and target relative elevation matrix are respectively the target relative position matrix, target relative azimuth matrix, and target relative elevation matrix at the moment after the state estimation is performed using the oth candidate data interconnection situation matrix as the final data interconnection mode.

Images