An optimization method for UAV leader-officer collaborative coverage based on ellipse fitting

The dual-UAV collaborative coverage method based on ellipse fitting solves the problems of low coverage efficiency and position fluctuation when there are many UAV targets with complex movements, and achieves efficient and stable target coverage.

CN115113651BActive Publication Date: 2024-11-08THE 54TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY GROUP CORPORATION
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210840950.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-07-18
Publication Date
2024-11-08
Estimated Expiration
2042-07-18

AI Technical Summary

Technical Problem

Existing UAV target coverage methods suffer from low optimization efficiency and fluctuations in optimal position when the number of targets far exceeds the number of UAVs and the target motion is difficult to analyze. In particular, effective coverage is difficult to achieve when dealing with grid boundaries.

Method used

A dual-UAV cooperative coverage method based on ellipse fitting is adopted. By performing minimum rectangle fitting on the target location point, the position of the ellipse focus is calculated, the UAV position is adjusted to cover the target, a safe distance is ensured, and the UAV position is updated in real time to adapt to the target movement.

Benefits of technology

It effectively reduces time costs, solves the problem of optimal position fluctuations caused by grid boundaries, and achieves efficient and stable target coverage, meeting the actual flight requirements of UAVs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115113651B_ABST
    Figure CN115113651B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for optimizing the coordinated coverage of unmanned aerial vehicles (UAVs) based on ellipse fitting, and belongs to the field of UAV control. The method includes the steps of setting the coverage range of the UAV and obtaining target parameters, performing ellipse fitting on the set of all target point positions, deploying the position of the UAV, updating the target position, and adjusting it online in real time. Compared with the traditional grid method or clustering method, this method adopts a suboptimal strategy in terms of the number of covered targets, thereby greatly reducing the time cost. Secondly, ellipse fitting is used instead of traditional clustering weighting, which effectively solves the optimal position fluctuation caused by the target near the grid boundary. When the target moves, the path of the UAV does not fluctuate too much, which is more in line with the actual flight requirements of the UAV. This method has very important practical significance for exploring the actual application mode of UAVs in various scenarios.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of cooperative control of unmanned aerial vehicles, and specifically relates to a solution for cooperative coverage optimization in a dual-unmanned aerial vehicle leader mode based on ellipse fitting. The application scenario thereof is mainly cooperative target coverage in which the number of targets is far greater than the number of unmanned aerial vehicles. Technical Background

[0002] Drones are widely used in all walks of life, including aerial photography, surveying and mapping, light shows, agricultural operations, logistics distribution, emergency rescue, etc. in the civil field, and reconnaissance, strikes, formations, air combat, relay communications, etc. in the military field. Among them, many drone application scenarios can be summarized as regional coverage issues. For example, in drone communication relay tasks in satellite communication blind spots, drones are required to cover as many ground communication users as possible; in group target monitoring tasks, drones are required to carry monitoring payloads to monitor as many targets as possible in the air; in electromagnetic interference tasks, drones are required to perform electromagnetic interference on as many ground communication devices as possible. The common point of these application scenarios is that the number of targets (communication users, monitoring targets, electromagnetic targets) is much greater than the number of drones, and the targets are not completely static, but are constantly moving, and the movement patterns are not easy to analyze, which greatly increases the difficulty of drones to provide optimal coverage of targets in real time.

[0003] There are many studies on the optimal coverage of drone targets at home and abroad. Common coverage strategies include cluster analysis, centroid extraction, density estimation, etc. However, due to the limitations of sensor detection range and drone flight time, it is difficult to achieve effective coverage in many cases. Considering that the essence of drone coverage is also an optimization problem, the focus is on mathematical modeling and planning algorithms. Traditionally, the mission area where the target is located is gridded, and the number of targets in each grid is counted. The coverage range of the drone is used as a window to move and traverse all gridded areas. If the drone is in a certain grid and covers the most targets, it is considered that this is the optimal coverage position of the drone. However, this method has two problems: 1) The optimization efficiency is low, and the entire target area needs to be traversed in the form of a sliding window, which is time-consuming; 2) When the target is moving, dealing with the grid boundary problem easily causes nonlinear fluctuations in the optimal position of the drone, making it difficult to actually operate the drone in specific flight. In order to solve the problem of high time cost, most of them consider increasing the number of drones to cover the task area in blocks. However, with the increase in the number of drones, the constraints between drones are also complicated. On the basis of the coverage problem, the problem of multi-machine coordination will be considered. At the same time, the grid boundary problem still causes the nonlinear fluctuation of the optimal result. Therefore, studying the key technologies and application limitations of drone coordination is of great practical significance for exploring the actual application mode of drones in various scenarios. Summary of the invention

[0004] In view of this, the present invention proposes a UAV chief-officer collaborative coverage optimization method based on ellipse fitting. The method is based on the premise that the two UAVs performing the search mission are isomorphic, that is, the two UAVs performing the coverage mission have the same performance, the same coverage sensors and coverage range; the number of targets is much greater than 2; in addition, the movement of the target must be a continuous motion mode, that is, the target cannot appear and disappear randomly, but only move randomly and continuously in the mission area, and the movement speed is less than the flight speed of the UAV, ensuring that the UAV can track and cover the target at any time.

[0005] The technical solution adopted by the present invention is:

[0006] A method for optimizing the coordinated coverage of unmanned aerial vehicles based on ellipse fitting includes the following steps:

[0007] Step 1: Setting the drone coverage area and obtaining target parameters

[0008] 1.1 Obtain the mission area to be covered, which is composed of polygons and is used to limit the execution range of the multi-machine collaborative coverage mission, and it is agreed that the target only moves within the mission area;

[0009] 1.2 Get the initial positions of the two drones, set as U1(m1,n1) and U2(m2,n2). The initial position of each drone is not limited to the mission area;

[0010] 1.3 Obtain the number of targets in the mission area and their current position distribution. Suppose the number of targets is N and the position of each target is T i (x i ,y i ) t , where i = 1, 2, ..., N, and t represents the current time;

[0011] 1.4 The effective coverage range of the payload of the drone is represented by a circle with a diameter of D, which is proportional to the flight altitude of the drone;

[0012] Step 2: Fit an ellipse to all target location points

[0013] 2.1 For all target positions T at time t i (x i ,y i ) t Perform minimum rectangle fitting to obtain the center point coordinates O(x0,y0), the short side length a, the long side length b, and the inclination angle θ with the x-axis of the minimum rectangle;

[0014] 2.2 Take the center point O(x0,y0) of the smallest rectangle as the center of the ellipse, a as the minor axis, and b as the major axis, and calculate the elliptic curve;

[0015] 2.3 Use the following formula to calculate the two foci P1 and P2 of the ellipse:

[0016]

[0017]

[0018] Step 3: Next position deployment of two drones

[0019] 3.1 Assume the safe distance between drones is L safe ;

[0020] 3.2 Determine the distance L between the two foci P1 and P2 calculated in step 2.3 p Is it less than L safe ;

[0021] 3.3 If L p Less than L safe , then connect P1 and P2, and make The extension line of the two drones is extended to points O1 and O2, and the two drones are deployed at points O1 and O2, so that the sum of the distances between the current positions of the two drones and the deployment positions is the shortest. If the sum of the distances in the two cases is the same, the deployment position is randomly selected;

[0022] 3.4 If L p Greater than or equal to L safe , then connect P1 and P2, and directly deploy the two drones at points P1 and P2, so that the sum of the distances between the current positions of the two drones and the deployment positions is the shortest. If the sum of the distances in the two cases is the same, then randomly select the deployment position;

[0023] Step 4: Real-time online adjustment of the drone position with target position update

[0024] 4.1 Set the target position update rate Δt, that is, update the target position once every Δt time;

[0025] 4.2 The position of each target at time t+Δt is T i (x i ,y i ) t+Δt , where i = 1, 2, ..., N;

[0026] Determine whether the end-mission instruction has been received; if received, the mission is ended; if not, return to step 2.2 to continue the ellipse fitting calculation, and obtain the new UAV deployment position, fly the UAV to the new position, and repeat this process until the end-mission command is received.

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

[0028] 1. The present invention adopts a dual-UAV cooperation mode. To solve the coverage problem of targets that are constantly moving and whose number is far greater than the number of UAVs, an ellipse fitting method is used to sparsely include the targets. The two foci of the fitted ellipse are used as the optimal coverage positions to deploy the positions of the dual UAVs. When the target moves, the major axis, minor axis and focus of the fitted ellipse will also change accordingly. According to the real-time change results of the focus, the deployment positions of the dual UAVs are continuously adjusted until the mission is completed.

[0029] 2. Compared with the traditional grid method or clustering method, this method adopts a suboptimal strategy in terms of the number of covered targets, which greatly reduces the time cost. Secondly, it uses ellipse fitting instead of traditional clustering weighting, which effectively solves the optimal position fluctuation caused by the target near the grid boundary. When the target moves, the path of the drone does not fluctuate too much, which is more in line with the actual flight needs of the drone. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the dual UAV collaborative coverage optimization method based on ellipse fitting.

[0031] Figure 2 Schematic diagram of the dual UAV collaborative coverage mission mode.

[0032] Figure 3 Schematic diagram of dual UAV collaborative coverage execution based on ellipse fitting.

[0033] Figure 4 Schematic diagram of simulation results. DETAILED DESCRIPTION

[0034] The present invention will be further described below in conjunction with the accompanying drawings.

[0035] A method for optimizing the coordinated coverage of unmanned aerial vehicles based on ellipse fitting includes the following steps:

[0036] Step 1: Setting the drone coverage area and obtaining target parameters

[0037] 1.1 Obtain the mission area to be covered. The mission area is generally composed of polygons, which is intended to limit the execution range of the multi-machine collaborative coverage mission, and stipulate that the target only moves within the mission area;

[0038] 1.2 Get the initial positions of the two drones, set as U1(m1,n1) and U2(m2,n2), and stipulate that the initial position of each drone can be set within the mission area or outside the mission area;

[0039] 1.3 Obtain the number of targets in the mission area and their current position distribution. Suppose the number of targets is N and the position of each target is T i(x i ,y i ) t , where i = 1, 2, ..., N, and t means that the position is the position of each target at time t;

[0040] 1.4 Set the effective coverage range of the payload for the drone to perform area coverage. The effective coverage range is represented by a circle with a diameter of D, which is usually proportional to the flight altitude of the drone.

[0041] Step 2: Fit an ellipse to all target location points

[0042] 2.1 For all target positions T at time t i (x i ,y i ) t Perform minimum rectangle fitting to obtain the center point coordinates O(x0,y0), the short side length a, the long side length b and the inclination angle θ with the x-axis of the minimum rectangle;

[0043] 2.2 Take the center point O(x0,y0) of the smallest rectangle as the center of the ellipse, a as the minor axis, and b as the major axis to calculate the elliptic curve;

[0044] 2.3 Use the following formula to calculate the two foci P1 and P2 of the ellipse:

[0045]

[0046]

[0047] Step 3: Next position deployment of two drones

[0048] 3.1 Assume the safe distance between drones is L safe ;

[0049] 3.2 Calculate the distance L between the two foci P1 and P2 calculated in step 2.3 p Is it less than L safe ;

[0050] 3.3 If L p Less than L safe , then connect P1 and P2, and make The extension line of line U1 O1 to points O1 and O2 is used to deploy the two drones at points O1 and O2, so that the sum of the distances between the current positions of the two drones and the deployment positions is the shortest. That is, if the length of line segment U1 O1 plus line segment U2 O2 is greater than the length of line segment U1 O2 plus line segment U2 O1, drone U1 is deployed at position O2 and drone U2 is deployed at position O1; if the length of line segment U1 O1 plus line segment U2 O2 is less than the length of line segment U1O2 plus line segment U2 O1, drone U1 is deployed at position O1 and drone U2 is deployed at position O2. If the distances in the two cases are the same, the deployment relationship is randomly selected.

[0051] 3.4 If L p Greater than or equal to L safe , then connect P1 and P2, and directly deploy the two UAVs at points P1 and P2, so that the sum of the distances between the current positions of the two UAVs and the deployment positions is the shortest. That is, if the length of line segment U1 P1 plus line segment U2 P2 is greater than the length of line segment U1 P2 plus line segment U2 P1, UAV U1 is deployed at position P2, and UAV U2 is deployed at position P1; if the length of line segment U1 P1 plus line segment U2 P2 is less than the length of line segment U1 P2 plus line segment U2 P1, UAV U1 is deployed at position P1, and UAV U2 is deployed at position P2. If the distances in the two cases are the same, the deployment relationship is randomly selected.

[0052] Step 4: Real-time online adjustment of the drone position with target position update

[0053] 4.1 Set the target position update rate Δt, that is, the target position is updated once every Δt time;

[0054] 4.2 Then the position of each target at time t+Δt is T i (x i ,y i ) t+Δt , where i = 1, 2, ..., N;

[0055] 4.3 Determine whether the end-mission instruction has been received; if received, the mission is ended; if not, return to step 2.2 to continue the ellipse fitting calculation, and obtain a new UAV deployment position, fly the UAV to the new position, and repeat this process until the mission end command is received.

[0056] Here is a more specific example:

[0057] A method for optimizing the coordinated coverage of UAV leaders based on ellipse fitting. The method flow is as follows: Figure 1 As shown, the specific steps include:

[0058] Step 1: Setting the drone coverage area and obtaining target parameters

[0059] 1.1 Obtain the mission area to be covered. The mission area is generally composed of polygons, which is intended to limit the execution range of the multi-machine collaborative coverage mission and stipulate that the target only moves within the mission area, such as Figure 2 As shown, the task area can be set as a rectangle Ω, the motion range of all targets is bounded by the entire rectangle, and the four locations of the rectangle Ω are AEFH;

[0060] 1.2 The initial positions of the two drones can be obtained in real time through the GPS module carried by each drone, which are set as U1(m1,n1) and U2(m2,n2). The initial position of each drone can be set within the mission area or outside the mission area;

[0061] 1.3 The number of targets and their current positions in the mission area are obtained in real time through the target detection payload on the drone. Suppose the number of targets is N and the position of each target is T. i (x i ,y i ) t , where i = 1, 2, ..., N, and t represents the position of each target at time t;

[0062] 1.4 The effective range of the payload covered by the drone is set as a circle with a diameter of D, which is generally proportional to the flight altitude of the drone. For example, if the effective coverage angle of the payload is α and the flight altitude of the drone is h, the coverage diameter is calculated by the following formula:

[0063] Step 2: Fit an ellipse to all target location points

[0064] 2.1 Perform minimum rectangle fitting on all target points, such as Figure 3 As shown in (a), the least square method is first used to fit the central axis of the minimum rectangle for all target points. Then, a straight line parallel to the central axis is constructed at the target position a that is the farthest from the central axis, and a straight line parallel to the central axis is constructed on the other side of the central axis at the same distance. At the two edge target positions b and c in the direction parallel to the central axis, two straight lines perpendicular to the central axis are constructed to form the minimum rectangle, as shown in Fig. Figure 3 (b), and then obtain the center point coordinates O(x0, y0), the length of the long side L, the length of the short side W and the inclination angle θ with the horizontal x-axis of the minimum rectangle;

[0065] 2.2 Take the center point O(x0,y0) of the smallest rectangle as the center of the ellipse, L as the length of the major axis, and W as the length of the minor axis, and calculate the elliptic curve, such as Figure 3 (c), the calculated elliptic curve equation is:

[0066]

[0067] in,

[0068] 2.3 Use the following formula to obtain the two foci P1 and P2 of the ellipse:

[0069]

[0070]

[0071] Step 3: Next position deployment of two drones

[0072] 3.1 Set the safe distance between the two drones to L safe , if L safe =D / 4;

[0073] 3.2 Calculate the distance L between the two foci P1 and P2 of the fitted ellipse P Is it less than L safe ;

[0074] 3.3 If L P Greater than or equal to L safe , then connect P1 and P2, and directly deploy the two drones at points P1 and P2. The sum of the distances between the current positions of the two drones and the deployment positions is the shortest, which is used as the decision basis for deploying drones U1 and U2 at P1 and P2 or P2 and P1 respectively. If the distances are the same, the deployment relationship is randomly selected, such as Figure 3 (d);

[0075] 3.4 If L P Less than L safe , then connect P1 and P2, and make The extension line to points O1 and O2, such as Figure 3 As shown in (e), the position of the Changli UAV is deployed at points O1 and O2. The sum of the distances between the current positions of the two UAVs and the deployment positions is the shortest, which is used as the decision basis for deploying UAVs U1 and U2 at O1 and O2 or O2 and O1 respectively. If the distances are the same, the deployment relationship is randomly selected, as shown in Figure 3 (f)

[0076] Step 4: Real-time online adjustment of the drone position with target position update

[0077] 4.3 Set the target position update rate Δt, that is, the target position is updated once every Δt time;

[0078] 4.4 Then the position of each target at time t+Δt is T i (x i ,y i) t+Δt , where i = 1, 2, ..., N;

[0079] 4.5 Determine whether the end-mission instruction has been received; if received, the mission is ended; if not, return to step 2.2 to continue the ellipse fitting calculation, and obtain a new UAV deployment position, fly the UAV to the new position, and repeat this process until the end-mission instruction is received.

[0080] Compared with the traditional grid method or clustering method, this method adopts a suboptimal strategy in terms of the number of covered targets, thereby greatly reducing the time cost. Secondly, it uses ellipse fitting instead of traditional clustering weighting to effectively solve the optimal position fluctuation caused by the target near the grid boundary. When the target moves, the path of the UAV does not fluctuate too much, which is more in line with the actual flight needs of the UAV.

[0081] The effect of this method is verified by simulation as follows:

[0082] In this embodiment, QT+VS2017 is used to build a simulation environment. The mission area is set as a quadrilateral, the number of drones is 2, the detection range of the drone is D=200 pixels, the flight speed of the drone is 20 pixels / second, the minimum safety distance between drones is 40 pixels, the initial positions of the two drones are U1=(500,300) and U1=(500,200), the number of targets is 100, and the targets are set to move irregularly in random directions through the simulation environment. The deployment positions of the two drones can be calculated based on the real-time distribution of the targets. The collaborative coverage effect of the two drones on 100 targets at different times is shown in Figure 2. Figure 4 As shown, sub-graphs ①-④ respectively represent the coverage of the entire target group by two drones under different target distribution conditions. This embodiment shows that the present invention can efficiently achieve the coverage of large-scale targets far exceeding the number of the drone itself, while ensuring the minimum safe distance between drones. This method has very important practical significance for exploring the actual application mode of drones in various scenarios.

[0083] The above description is only a specific example of the present invention and does not constitute any limitation to the present invention. It is obvious that for professionals in this field, after understanding the content and principle of the present invention, various modifications and changes in form and details may be made without departing from the principle and structure of the present invention, but these modifications and changes based on the idea of ​​the present invention are still within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing the coordinated coverage of unmanned aerial vehicles based on ellipse fitting, characterized in that it includes the following steps: Step 1: Setting the drone coverage area and obtaining target parameters 1.1 Obtain the mission area to be covered, which is composed of polygons and is used to limit the execution range of the multi-machine collaborative coverage mission, and it is agreed that the target only moves within the mission area; 1.2 Get the initial positions of the two drones, set as U1(m1,n1) and U2(m2,n2). The initial position of each drone is not limited to the mission area; 1.3 Obtain the number of targets in the mission area and their current position distribution. Suppose the number of targets is N and the position of each target is T i (x i ,y i ) t , where i = 1, 2, ..., N, and t represents the current time; 1.4 The effective coverage range of the payload of the drone is represented by a circle with a diameter of D, which is proportional to the flight altitude of the drone; Step 2: Fit an ellipse to all target location points 2.1 For all target positions T at time t i (x i ,y i ) t Perform minimum rectangle fitting to obtain the center point coordinates O(x0,y0), the short side length a, the long side length b, and the inclination angle θ with the x-axis of the minimum rectangle; 2.2 Take the center point O(x0,y0) of the smallest rectangle as the center of the ellipse, a as the minor axis, and b as the major axis, and calculate the elliptic curve; 2.3 Use the following formula to calculate the two foci P1 and P2 of the ellipse: Step 3: Next position deployment of two drones 3.1 Assume the safe distance between drones is L safe ; 3.2 Determine the distance L between the two foci P1 and P2 calculated in step 2.3 p Is it less than L safe ; 3.3 If L p Less than L safe , then connect P1 and P2, and make The extension line of the two drones is extended to points O1 and O2, and the two drones are deployed at points O1 and O2, so that the sum of the distances between the current positions of the two drones and the deployment positions is the shortest. If the sum of the distances in the two cases is the same, the deployment position is randomly selected; 3.4 If L p Greater than or equal to L safe , then connect P1 and P2, and directly deploy the two drones at points P1 and P2, so that the sum of the distances between the current positions of the two drones and the deployment positions is the shortest. If the sum of the distances in the two cases is the same, then randomly select the deployment position; Step 4: Real-time online adjustment of the drone position with target position update 4.1 Set the target position update rate Δt, that is, update the target position once every Δt time; 4.2 The position of each target at time t+Δt is T i (x i ,y i ) t+Δt ,in, i=1,2,…,N; 4.3 Determine whether the end-task instruction has been received; if received, the task is ended; If not, go back to step 2.2 to continue the ellipse fitting calculation, and find the new UAV deployment position, fly the UAV to the new position, and repeat this process until the mission end command is received.