Unmanned aerial vehicle cooperative optimal coverage detection method for aerial cluster target

Abstract

The invention discloses an unmanned aerial vehicle cooperative optimal coverage detection method for an aerial cluster target. The method comprises the following steps: establishing a target probability distribution model and an unmanned aerial vehicle detection model under radar detection; based on the target probability part model and the unmanned aerial vehicle detection model, establishing a cluster collaborative coverage detection optimization target function; and according to real-time detection results of the radar and the unmanned aerial vehicle, iterative optimization is carried out on the optimization target function based on the detection results, and coverage detection of the target group is realized. According to the unmanned aerial vehicle cooperative optimal coverage detection method for the aerial cluster target disclosed by the invention, under multi-error input, an expected unmanned aerial vehicle formation and a sight line direction can be quickly generated, and high-coverage-rate detection of a target cluster is realized.

Description

Technical Field

[0001] The invention relates to a drone coordinated optimal coverage detection method for aerial cluster targets, and belongs to the technical field of aircraft control. Background Art

[0002] In many cases, drone swarms are needed to coordinate area coverage detection.

[0003] Regarding the problem of cooperative detection of UAVs, most existing technologies are based on ideal assumptions and do not take into account engineering constraints such as information errors, control errors, and quantity limitations that exist in reality. In addition, most algorithms are time-consuming and difficult to apply to highly dynamic, practical cluster confrontations.

[0004] For example, the prior art discloses an area coverage detection algorithm that relies on clustering of the time window of the target area, and then uses each cluster and the prepared drone increments to generate multi-unmanned and collaborative trajectory planning problems, and then solves the trajectory planning problem by using the simplex algorithm. However, this algorithm is based on a certain initial formation of drones to achieve coverage detection of static areas through task division and trajectory planning, and is not suitable for real-time coverage detection tasks in dynamic combat scenarios.

[0005] For example, the prior art also includes linear programming models based on ant colony algorithms and hexagonal grids to cover different areas. Based on the actual UAV performance, the UAV area coverage search trajectory is generated with time optimization as the guide, which improves the coverage search efficiency of small UAV clusters. However, this type of algorithm can only be applied to scenarios with fixed target areas, and is not applicable when covering and detecting high-speed moving target clusters.

[0006] Therefore, it is necessary to conduct more in-depth research on the collaborative coverage detection method of drone clusters to solve the above problems. Summary of the invention

[0007] In order to overcome the above problems, the inventors conducted in-depth research and proposed a UAV collaborative optimal coverage detection method for aerial cluster targets, including the following steps:

[0008] S1. Establish target probability distribution model and UAV detection model under radar detection;

[0009] S2, establishing cluster collaborative coverage detection optimization objective function based on the target probability partial model and UAV detection model;

[0010] S3. According to the real-time detection results of radar and UAV, the optimization objective function is iteratively optimized based on the detection results to achieve coverage detection of the target group.

[0011] Preferably, in S1, the target probability distribution model under radar detection is a planar two-dimensional model.

[0012] Preferably, in S1, the target probability distribution model under radar detection is expressed as:

[0013]

[0014]

[0015] Among them, P kk It means that after the plane area is discretized into a grid, any point in the mth grid has coordinates x k ,y k ; represents the probability of the target appearing at this point, j is the target number, n represents the total number of targets in the target cluster, T j represents the jth target, Indicates the target T j Appears at point P kk (x k ,y k ), Indicates the target T detected by radar j The position of x ,σ y Indicates the error between the target position detected by the radar and the actual position on the plane.

[0016] Preferably, σ x ,σ y Satisfies the normal distribution standard, expressed as:

[0017] σ x =σ D cos(α E )*cos(α A )+D xy σ Az sin(α A )+Dσ EL sin(α E )*cos(α A )

[0018] σ y =σ D cos(α E )*sin(α A )+D xy σ AZ cos(α A )+Dσ EL sin(α E )*sin(α A )

[0019] Among them, σ D represents the radar’s distance detection error, σAZ represents the azimuth detection error of the radar, σ EL Represents the radar’s elevation angle detection error, α A represents the azimuth of the radar detecting the target, α E Indicates the elevation angle of the radar detecting the target.

[0020] Preferably, the drone detection model is expressed as:

[0021]

[0022] in, Indicates drone U i Detection point P kk The probability of Indicates drone U i With point P kk The distance, R represents the distance of the drone U i The field of view radius, Indicates point P kk With UAV i The angle, α 1 , α 2 , α 3 is the configurable field of view parameter, k 1 , k 2 , k 3 is a configurable probability parameter.

[0023] Preferably, in S2, the optimization objective function is set to:

[0024]

[0025] Among them, Reward(U) is the optimization target, i is the number of the drone, K i is the distance function of the ith UAV, ω 1 ,ω 2 ,ω 3 ,ω 4 is a configurable parameter, d represents the distance between the ith UAV and the nearest UAV, Reword i For UAV i The total detection benefit is , N represents the total number of UAVs, M represents the total number of discrete grids, M = k × k, k is the number of divisions of a single-sided grid, and m is the number of the grid.

[0026] Preferably, in S3, the gray wolf optimization algorithm is used to iteratively optimize the optimization objective function.

[0027] The beneficial effects of the present invention include:

[0028] (1) Under multiple error inputs, it can quickly generate the expected UAV formation and line of sight direction, achieving high coverage detection of the target cluster;

[0029] (2) It meets the flight safety constraints of UAVs and can meet the needs of high-dynamic coverage detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A schematic diagram of a flow chart of a method for cooperative optimal coverage detection of aerial cluster targets by unmanned aerial vehicles according to a preferred embodiment of the present invention is shown;

[0031] Figure 2 It shows that in Example 1, the target groups are arranged in rectangular, triangular and circular arrays, respectively, and the coverage detection results are iteratively optimized;

[0032] Figure 3 The relationship between the number of iterations and the computation time is shown in Example 1 when the target groups are arranged in a triangular array. DETAILED DESCRIPTION

[0033] The present invention will be further described in detail below through the accompanying drawings and embodiments. Through these descriptions, the characteristics and advantages of the present invention will become more clear and distinct.

[0034] The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration." Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments. Although various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless otherwise noted.

[0035] According to the present invention, a UAV cooperative optimal coverage detection method for aerial cluster targets is provided. Figure 1 As shown, the following steps are included:

[0036] S1. Establish target probability distribution model and UAV detection model under radar detection;

[0037] S2, establishing cluster collaborative coverage detection optimization objective function based on the target probability partial model and UAV detection model;

[0038] S3. According to the real-time detection results of radar and UAV, the optimization objective function is iteratively optimized based on the detection results to achieve coverage detection of the target group.

[0039] In S1, the target probability distribution model under radar detection is a two-dimensional plane model. Most of the formations of drone clusters do not have a large height difference and can be detected by the field of view of the drone in the height direction. Therefore, the target probability distribution model can obtain a relatively good effect by using a two-dimensional plane model.

[0040] Preferably, in S1, the target probability distribution model under radar detection is expressed as:

[0041]

[0042] Among them, P kk It means that after the plane area is discretized into a grid, any point in the grid has its coordinate x k ,y k ; represents the probability of the target appearing at this point, j is the target number, n represents the total number of targets in the target cluster, T j represents the jth target, Indicates the target T j Appears at point P kk (x k ,y k ), Indicates the target T detected by radar j The position of x ,σ y Indicates the error between the target position detected by the radar and the actual position on the plane.

[0043] Preferably, σ x ,σ y Satisfies the normal distribution standard, expressed as:

[0044] σ x =σ D cos(α E )*cos(α A )+D xy σ AZ sin(α A )+Dσ EL sin(α E )*cos(α A )

[0045] σ y =σ D cos(α E )*sin(α A )+D xy σ AZ cos(α A )+Dσ EL sin(α E )*sin(α A )

[0046] Among them, σ D represents the radar’s distance detection error, σ AZ represents the azimuth detection error of the radar, σ EL Represents the radar’s elevation angle detection error, α Arepresents the azimuth of the radar detecting the target, α E Indicates the elevation angle of the radar detecting the target.

[0047] The UAV detection model is expressed as:

[0048]

[0049] in, Indicates drone U i Detection point P kk The probability of Indicates drone U i With point P kk The distance, R represents the distance of the drone U i The field of view radius, Indicates point P kk With UAV i The angle, α 1 , α 2 , α 3 is the configurable field of view parameter, k 1 , k 2 , k 3 is a configurable probability parameter.

[0050] The inventor found that in the actual process of performing detection tasks, the detection effect of the optoelectronic pod is related to the angle between the target and the direction of the drone's line of sight. Targets that appear in the middle area of ​​the field of view are often easier to be stably identified and tracked, while the effect is slightly worse in the peripheral areas within the field of view. Because of the existence of the control pointing error of the pod, unexpected targets sometimes appear at the edge of the field of view, and sometimes targets that should have appeared in the field of view are sometimes lost. In the present invention, the field of view area is divided into three areas: the central area, the peripheral area, and the uncertain detection area through the drone detection model. Whether the target is in the visible area is determined by the distance and the angle between the target and the line of sight, and different detectable probability benefit values ​​are assigned to the target according to the size of the angle. Only when the line of sight angle of the target appears in the field of view, that is, the center and peripheral areas, is it considered to be seen. Although there is also benefit in the possible detection area, it is not considered to have been detected and can be detected repeatedly.

[0051] In S2, the optimization objective function is set to:

[0052]

[0053] Among them, Reward(U) is the optimization target, i is the number of the drone, K i is the i-th UAV spacing function, which is used to ensure the safe distance between different UAVs in the UAV cluster, ω 1 ,ω 2 ,ω3 ,ω 4 is a configurable parameter, d represents the distance between the ith UAV and the nearest UAV, Reword i For UAV i The total detection benefit is , N represents the total number of UAVs, M represents the total number of discrete grids, m = 1, 2, ..., M = k × k, k is the number of divisions of a single-sided grid, and m is the number of the grid.

[0054] According to the present invention, the drone set is U{U 1 ,U 2 ,…U i …U N}, and each drone has four parameters: radius R, field of view angle α, line of sight direction θ and position coordinates (x, y).

[0055] According to the present invention, the optimization objective function takes maximizing detection gain as the optimization goal, and takes the UAV position and line of sight direction as the optimization decision variables. It fully considers the low detection accuracy constraint of radar for small group targets, the limited field of view constraint of UAV, and the detection characteristics of visual recognition algorithm, and generates our UAV formation and expected line of sight direction in real time according to the error-containing input of radar, so as to achieve high coverage detection of target clusters.

[0056] Preferably, in S3, the gray wolf optimization algorithm is used to iteratively optimize the optimization objective function. The gray wolf algorithm has the characteristics of simple structure, easy implementation and high solution efficiency.

[0057] The Gray Wolf Optimization Algorithm is a new type of swarm intelligence algorithm proposed in 2015. It is an optimization algorithm based on the hunting behavior of wolf packs. For details, please refer to the literature

[22] .

[0058] Rui W,Shiyao H,Wenqian Z,et al.An improved grey wolf optimizer forminimizing drilling deformation and residual stress in AA2024 sheet[J].TheInternational Journal of Advanced Manufacturing Technology,2024,130(9-10):4443-4458.It is not described in detail in the present invention.

[0059] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the disclosure of the present invention can be performed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and this document does not limit this.

[0060] Example

[0061] Example 1

[0062] Conduct simulation experiments to perform drone cluster collaborative coverage detection, including the following steps:

[0063] S1. Establish target probability distribution model and UAV detection model under radar detection;

[0064] S2, establishing cluster collaborative coverage detection optimization objective function based on the target probability partial model and UAV detection model;

[0065] S3. According to the real-time detection results of radar and UAV, the optimization objective function is iteratively optimized to achieve coverage detection of the target group.

[0066] In S1, the target probability distribution model under the radar detection is expressed as:

[0067]

[0068] σ x ,σ y Satisfies the normal distribution criteria, expressed as:

[0069] σ x =σ D cos(α E )*cos(α A )+D xy σ AZ sin(α A )+Dσ EL sin(α E )*cos(α A )

[0070] σ y =σ D cos(α E )*sin(α A )+D xy σ AZ cos(α A )+Dσ EL sin(α E )*sin(α A )

[0071] The UAV detection model is expressed as:

[0072]

[0073] In S2, the optimization objective function is set to:

[0074]

[0075] In S3, the gray wolf optimization algorithm is used to iteratively optimize the optimization objective function. The number of wolf packs in the gray wolf optimization algorithm is set to 30, and the number of algorithm iterations is 10.

[0076] During the simulation, the number of drones in the drone swarm is set to 5, the drone field of view angle is 7°, the drone field of view angle range is 1000m, and the field of view can be adjusted from -20° to 20°.

[0077] Set the radar and drone error parameters as shown in Table 1:

[0078] Table 1

[0079]

[0080]

[0081] The parameters involved in the simulation process are shown in Table 2:

[0082] Table 2

[0083]

[0084] Set the target group to be arranged in rectangular, triangular, and circular arrays, and iteratively optimize the coverage detection results. Figure 2 As shown in the figure, for different target formations, an average coverage rate of more than 98% can be achieved after 8 iterations, and the standard deviation of the coverage rate is also below 8%, which can basically guarantee a coverage rate of more than 90%. If iterates 10 times, the average coverage rate has reached more than 99.4%, and the standard deviation has also been reduced to less than 3.5%, which can stably achieve a coverage rate of more than 95%.

[0085] The onboard computer performs iterative optimization. The target group is arranged in a triangular array. The relationship between the number of iterations and the operation time is recorded. The results are as follows: Figure 3 As shown in the figure, it can be seen that under the condition of achieving high probability coverage of the target cluster, 10 optimization iterations can be basically completed within 1.2S to achieve high coverage, that is, dynamic real-time detection can be realized.

[0086] The present invention has been described above in conjunction with preferred embodiments, but these embodiments are only exemplary and serve only as an illustration. On this basis, the present invention may be subjected to a variety of substitutions and improvements, all of which fall within the scope of protection of the present invention.

Claims

1. A UAV cooperative optimal coverage detection method for aerial cluster targets, characterized in that: The following steps are involved: S1. Establish target probability distribution model and UAV detection model under radar detection; S2, establishing cluster collaborative coverage detection optimization objective function based on the target probability partial model and UAV detection model; S3. According to the real-time detection results of radar and UAV, the optimization objective function is iteratively optimized based on the detection results to achieve coverage detection of the target group.

2. The method for detecting drone-assisted optimal coverage of aerial cluster targets according to claim 1 is characterized in that: In S1, the target probability distribution model under radar detection is a planar two-dimensional model.

3. The method for detecting drone-coordinated optimal coverage of aerial cluster targets according to claim 1 is characterized in that: In S1, the target probability distribution model under the radar detection is expressed as: Among them, P kk It means that after the plane area is discretized into a grid, any point in the grid has its coordinate x k ,y k ; represents the probability of the target appearing at this point, j is the target number, n represents the total number of targets in the target cluster, T j represents the jth target, Indicates the target T j Appears at point P kk (x k ,y k ), Indicates the target T detected by radar j The position of x , σ y Indicates the error between the target position detected by the radar and the actual position on the plane.

4. The method for detecting drone-assisted optimal coverage of aerial cluster targets according to claim 3 is characterized in that: σ x , σ y Satisfies the normal distribution criteria, expressed as: s x =s D cos(a E )*cos(α A )+D xy s AZ sin(a A )+Dσ EL sin(a E )*cos(α A ) s y =s D cos(a E )*sin(a A )+D xy s AZ cos(a A )+Dσ EL sin(a E )*sin(a A ) Among them, σ D represents the radar’s distance detection error, σ AZ represents the azimuth detection error of the radar, σ EL Represents the radar’s elevation angle detection error, α A represents the azimuth of the radar detecting the target, α E Indicates the elevation angle of the radar detecting the target.

5. The method for detecting drone-coordinated optimal coverage of aerial cluster targets according to claim 1 is characterized in that: The UAV detection model is expressed as: in, Indicates drone U i Detection point P kk The probability of Indicates drone U i With point P kk The distance, R represents the distance of the drone U i The field of view radius, Indicates point P kk With UAV i The angles formed, α1, α2, α3 are configurable field angle parameters, and k1, k2, k3 are configurable probability parameters.

6. The method for detecting drone-assisted optimal coverage of aerial cluster targets according to claim 1 is characterized in that: In S2, the optimization objective function is set to: Among them, Reward(U) is the optimization target, i is the number of the drone, K i is the distance function of the ith UAV, ω1, ω2, ω3, ω4 are configurable parameters, d represents the distance between the ith UAV and the nearest UAV, Reword i For UAV i The total detection benefit is , N represents the total number of UAVs, M represents the total number of discrete grids, and m is the number of the grid.

7. The method for detecting drone-assisted optimal coverage of aerial cluster targets according to claim 1, characterized in that: In S 3, the gray wolf optimization algorithm is used to iteratively optimize the optimization objective function.

Images