Multi-unmanned aerial vehicle cooperative hunting method in obstacle environment

Abstract

The invention belongs to the technical field of multi-quad-rotor collaborative hunting, and discloses a multi-unmanned aerial vehicle collaborative hunting method in an obstacle environment, which comprises the following steps: step 1, three-dimensional decision space representation based on a Voronoi diagram: simplified representation of a three-dimensional obstacle decision space is realized by constructing a safe priority reachable area; and step 2, constructing a distributed hunting decision based on area minimization: on the basis of a security priority reachable region constructed by a three-dimensional decision space representation module based on a Voronoi diagram, compressing a motion region of the escape unmanned aerial vehicle by adopting a collaborative hunting strategy based on area minimization, and finally realizing effective capture. According to the three-dimensional decision space characterization method based on the Voronoi diagram, the convex bounded decision area can be extracted from the three-dimensional obstacle space which is a typical non-convex area, effective characterization of the decision space is achieved, and a foundation is laid for generation of a subsequent encircle strategy.

Description

Technical Field

[0001] The present invention belongs to the technical field of multi-quadrotor cooperative capture, and in particular relates to a multi-UAV cooperative capture method in an obstacle environment. Background Art

[0002] Multi-quadrotor collaborative capture missions usually involve multiple quadrotors coordinating to track and surround one or more escaping targets to prevent them from escaping. This type of task belongs to the multi-robot pursuit and escape (MPE) game, which is widely used in areas such as area search, monitoring, and target tracking. Existing solutions mainly rely on solving differential equations, such as the differential game method based on the Hamilton-Jacobi-Isaacs (HJI) equation. Although this method can provide the optimal control strategy in complex environments, it is difficult to handle large-scale problems due to its high dimensionality and computational complexity of the solution. In addition, although the reinforcement learning method can train collaborative pursuit strategies, its computational complexity is high and its adaptability to the environment is poor. Summary of the invention

[0003] The purpose of the present invention is to provide a multi-UAV collaborative capture method in an obstacle environment to solve the above-mentioned technical problems.

[0004] In order to solve the above technical problems, the specific technical solution of a multi-UAV cooperative capture method in an obstacle environment of the present invention is as follows:

[0005] A method for cooperatively capturing multiple drones in an obstacle environment, which obtains spatial coordinate information of multiple drones including a defensive drone and an offensive drone according to relevant position capture equipment and algorithms, includes the following steps:

[0006] Step 1: Three-dimensional decision space representation based on Voronoi diagram: A simplified representation of the three-dimensional obstacle decision space is achieved by constructing a safe priority accessible area;

[0007] Step 2: Construct a distributed capture decision based on area minimization: Based on the safe priority reachable area constructed by the three-dimensional decision space representation module based on the Voronoi diagram, a collaborative capture strategy based on area minimization is adopted to compress the movement area of ​​the escaping drone and ultimately achieve effective capture.

[0008] Furthermore, the step 1 comprises the following steps:

[0009] Step 1.1: Construct a priority reachable domain based on the Voronoi diagram: Depending on the location of the capture drone cluster and the location of the escape drone, the Voronoi partition method is used in the three-dimensional scene to determine the priority reachable area of ​​the capture drone in the three-dimensional space.

[0010] Step 1.2: Construction of safe priority reachable domain based on obstacle perception: Based on the priority reachable area, the safe distance between the captured drones and the minimum distance between the captured drones and obstacles are considered. By constructing a safe priority reachable domain, it is ensured that these constraints are always met during the entire capture process.

[0011] Step 1.3: Area iteration and update: Based on real-time scenario changes, continuously adjust the safe priority accessible areas to lay the foundation for high-frequency real-time re-decision-making.

[0012] Furthermore, the step 1.1 includes the following specific steps:

[0013] The escapee's priority accessible area is defined as the escapee's priority accessible area in a given environment. The set of all locations that can be reached faster than any other roundup;

[0014] Given a set of seed points P = {p1, p2, ..., p n}, the Voronoi diagram divides the space into n convex polygonal regions, called Voronoi regions, or Dirichlet regions, each Voronoi region V(p i ) consists of all the distances from the seed point p i The nearest point is formed, satisfying the following conditions:

[0015]

[0016] Among them, Q is the task scene space, x is any point in the space, and p i is the seed point, and the point in the Voronoi region to the seed point p i The distance to is smaller than the distance to other seed points;

[0017] In the two-dimensional plane, each Voronoi region of the Voronoi diagram is a convex polygon, while in three-dimensional space, the region is composed of polyhedrons. The Voronoi edge in the Voronoi diagram is the shared boundary between adjacent seed points, which is defined as the set of all points with equal distances to two adjacent seed points. In the three-dimensional scene, the shared Voronoi boundary is represented as the perpendicular bisector of the line connecting the two seed points. Mathematically, a point x on the shared Voronoi boundary is g The following conditions are met:

[0018]

[0019] The priority accessible areas for escapers and the priority accessible areas for capturers are defined as:

[0020]

[0021] in, and They represent the priority accessible areas of the captors and escapees in an obstacle-free environment. In this environment, each agent can determine its accessible area in space by calculating the Voronoi diagram, thereby optimizing its path planning and control strategy.

[0022] The parameterized expression of the priority reachable domain is realized by introducing the concept of hyperplane:

[0023] The hyperplane in a three-dimensional environment can be defined by the normal vector and the coordinate representation of a point on the hyperplane. Assume that there is a known point P0 (x0, y0, z0) and a normal vector on the hyperplane. Then the equation of the hyperplane is expressed as:

[0024]

[0025] in, is an arbitrary point in space. The geometric object represented by this equation divides space into two half-spaces: one region satisfying n1(x-x0)+n2(y-y0)+n3(z-z0)>0, and the other region satisfying n1(x-x0)+n2(y-y0)+n3(z-z0)<0;

[0026] According to the hyperplane separation rule, the Voronoi partition of each capturer can be represented by a set of hyperplanes. The combination of the internal sub-regions of multiple hyperplanes can form a convex region, which represents the priority reachable area of ​​the capturer or escaper.

[0027] Therefore, the definition of the priority reachable region PRR can be transformed from formula (3) into the following form:

[0028]

[0029] Here, X represents the vector from the origin to any point x in Q, and l1 and l2 are used to refer to different agents including escapers and capturers. and are the parameters of the hyperplane defined by the difference in positions between agents l1 and l2, Represents the normal vector of the corresponding hyperplane and the plane position calculated by the midpoint;

[0030] Calculate the hyperplane normal vector and corresponding plane parameters associated with each pair of agents. The normal vector of the hyperplane is determined by the relative position difference between the capturer or escaper, and the position parameter of the plane It is calculated by the midpoint between the two, ensuring that the hyperplane separates the regions according to the definition of Voronoi partition. The specific calculation formula is as follows:

[0031]

[0032] In the above formula, and denote the position vectors of agents l1 and l2 respectively, and is the relative position vector between them, and the plane position parameter is determined by the midpoint of these two locations, ensuring that each hyperplane accurately separates the space.

[0033] Furthermore, the step 1.2 includes the following specific steps:

[0034] The safe flight area under the current position of the drone is represented by constructing a convex polyhedron, which is composed of multiple separating hyperplanes. Each hyperplane separates the drone from obstacles to form a safe and obstacle-free flight space. The mathematical representation of this space is as follows:

[0035]

[0036] in, Indicates the safe reachable area for drones. Defines the composition v l The parameter set of the hyperplane;

[0037] The KD-Tree data structure is used to speed up the retrieval process of obstacle points and separate the hyperplane parameters. The calculation formula is as follows:

[0038]

[0039] Among them, r s is the safety radius, which is used to consider the physical size of the drone and ensure the obstacle avoidance effect during flight by offsetting the hyperplane a certain distance away from the obstacle. By combining formula (5) and formula (7), the final safe priority reachable area is defined as follows:

[0040]

[0041] Furthermore, the step 1.3 includes the following specific steps:

[0042] Recalculate the safe priority accessible area based on the real-time changing location information,

[0043] 1) Input information: current location x of the capture drone i , i∈Γ, the current position of the target escaping drone x e , obstacle point cloud information Ω_obs in the environment;

[0044] 2) Initialize the remaining obstacle point cloud information Ω_remain=Ω_obs. Assign the obstacle point cloud information Ω_obs in the environment to the remaining obstacle point cloud information Ω_remain;

[0045] 3) Construct a separating hyperplane based on the obstacle point cloud;

[0046] Enter a loop, and as long as the number of points in the remaining obstacle point cloud information is greater than a preset threshold, perform the following operations:

[0047] 3-1) Find the nearest obstacle point: Use the KD tree algorithm to find the nearest obstacle point to the current capture drone position x i The nearest obstacle point o_near;

[0048] 3-2) Construct obstacle separation hyperplane: calculate parameters Use these parameters to construct an obstacle separation hyperplane and store the relevant information of the hyperplane in the safe priority reachable area middle;

[0049] 3-3) Update the remaining obstacle point cloud information: remove the points outside the just constructed hyperplane from the remaining obstacle point cloud information to prepare for the next cycle;

[0050] 4) Dealing with other drones

[0051] Through a loop to traverse all the captured drones, that is, j ranges from 0 to N and j is not equal to i:

[0052] 4-1) Set l1 as the number of the current captured drone, and l2 as the number of the currently traversed captured drone;

[0053] 4-2) Calculate parameters based on specific formulas <n ij , p ij > and store these parameters in the safe priority reachable area middle;

[0054] 5) Dealing with target escaping drones

[0055] Set l1 to the number of the current captured drone, l2 to the number of the target escaping drone, and calculate the parameters according to a specific formula <n ie , p ie > and store these parameters in the safe priority reachable area middle;

[0056] 6) After the above steps, the safe priority accessible area is finally obtained as the output of the algorithm.

[0057] Furthermore, the step 2 includes the following specific steps:

[0058] Step 2.1: Construct a collaborative capture strategy based on area minimization: Based on the safe priority reachable domain constructed in step 1.2, the escaped drone is effectively captured by reducing the safe priority reachable area of ​​the escaped drone;

[0059] Step 2.2: Prove the effectiveness of the round-up strategy: Prove the theoretical completeness of the collaborative round-up strategy based on area minimization proposed in step 2.1 through theoretical analysis;

[0060] Step 2.3: Distributed collaborative roundup decision framework: lightweight algorithm by allocating computational load.

[0061] Furthermore, the step 2.1 includes the following specific steps:

[0062] The area of ​​the escapee's safe priority accessible area is defined as:

[0063]

[0064] in is the safe priority accessible area of ​​the escapee constructed in formula (9). According to the definition of formula (10), The derivative of is:

[0065]

[0066] In formula (11), the capture of drones The influence of can be decoupled as Through this decoupling, the hunter adopts an "area minimization" movement strategy that is consistent with The gradient descent direction is consistent with that of , which can achieve a rapid reduction in the area of ​​the escapee's safe priority accessible area. The movement strategy of the capturer is given by the following formula:

[0067]

[0068] Derived the specific calculation method of the above strategy, using the Leibniz integration rule, can be simplified to:

[0069]

[0070] Among them, N s represents the set of capturing drones i that share a boundary with the escaping drone e, s i represents the shared boundary between the two. The area of ​​the shared boundary can be expressed as The centroid of the shared boundary is defined as

[0071] By comparing equation (11) and equation (13), we can get:

[0072]

[0073] Therefore, the area of ​​the decoupled safe priority reachable area of ​​the escaping drone is affected by the capturing drone as follows:

[0074]

[0075] Substituting equation (15) into equation (12), the final round-up movement strategy is simplified to:

[0076]

[0077] The "area minimization" collaborative capture strategy guides the capturer to the centroid of the shared boundary with the escapee. This strategy gradually reduces the area of ​​the escapee's safe priority reachable area, and eventually achieves the capture of the escapee.

[0078] Furthermore, the step 2.2 includes the following specific steps:

[0079] Lemma 2: Consider a mission scenario where a single capture drone and a single escape drone move in region Q. According to the proposed capture strategy (16), for any allowed escape strategy, Furthermore, if and only if the escaper adopts the following movement strategy,

[0080]

[0081] Among them, C s Is a tracking drone x i and escape dronex e The centroid of the shared boundary of the safety-first reachable area. Proof: For the case of a single pursuer and a single escaper, Simplified to:

[0082]

[0083] Substituting the area minimization-based round-up strategy (16) obtained in step 2.1 into the above formula, we obtain:

[0084]

[0085] Since the capture drone and the escape drone move at the same speed, the shared boundary between them is represented by the center of mass on the perpendicular bisector between them. is located on this shared boundary, so we have And without loss of generality in step 2.1, assume that v i =ve =1, that is Substituting the above conditions, we can see that for a single tracker, And when When , the control strategy of the escaper is:

[0086]

[0087] So far, Lemma 2 has been proved. The movement strategy of the hunter can ensure that the area of ​​the escapee's safe reachable area is strictly non-increasing, and the only escapee strategy that can keep the area of ​​the escapee's safe reachable area unchanged is to move in the opposite direction of the centroid of the shared boundary between the two.

[0088] The square of the distance between the escaper and the pursuer is defined as:

[0089] D=||x i -x e || 2 =(x i -x e ) T (x i -x e ) (twenty one)

[0090] Lemma 3: When the hunter adopts the movement strategy (16), the rate of change of the square of the distance satisfy In addition, if and only if the escaper adopts the control strategy shown in formula (20),

[0091] Proof: According to formula (21), the definition of D is

[0092]

[0093] Substituting the movement strategy (16) of the hunter into the above formula, we get:

[0094]

[0095] Also due to Located on the perpendicular bisector of the capture drone and the escape drone, so The above formula can be further deduced as:

[0096]

[0097] Without loss of generality in step 2.1, the present invention assumes that v i =v e =1, that is Therefore, there is and The motion strategy of the escaping drone is shown in formula (20);

[0098] So far, Lemma 3 has been proved. The movement strategy of the hunter can ensure that the distance between him and the escaper is strictly non-increasing. For the escaper, the only escape strategy that can keep the distance unchanged is to move in the opposite direction of the centroid of the shared boundary between the two.

[0099] According to Lemma 2 and Lemma 3, it is concluded that in the scenario of a single capture drone, when the capture drone adopts the motion strategy shown in formula (16), the area of ​​the escapee's safe priority reachable area and the distance of the escapee from the capture drone are strictly non-increasing, and the rate of change of the two is zero only when the escapee moves in the opposite direction of the shared boundary centroid of the two, and in other cases both are decreasing; therefore, returning to the multi-drone collaborative capture scenario described in the present invention, when the capture drone adopts the capture strategy described in formula (16), no matter what motion strategy the escapee adopts, it cannot guarantee that the condition that its distance from each adjacent capture drone is not increasing is met at the same time, so the distance between the escapee and the capture drone cluster will always be in a decreasing state, and eventually the capture drone cluster will be finally captured within a limited time.

[0100] Furthermore, the step 2.3 includes the following specific steps:

[0101] 1) Initialization

[0102] For each captured drone, initialize the following variables:

[0103] A set of hyperplane parameters used to store safe priority reachable areas;

[0104] vertices_set: used to store the vertex set of each face of the convex polyhedron formed by the safe priority reachable area;

[0105] shareplane: A flag used to indicate whether the current capture drone and the target share a boundary;

[0106] 2) Traverse each captured drone

[0107] Through a loop, the following operations are performed on each drone in the capture drone cluster: 2-1) Reset the shared boundary flag: Initially set the shareplane flag to True, indicating that it is assumed that the current capture drone has a shared boundary with the target;

[0108] 2-2) Constructing a safe priority reachable area: Call the SPRRConstruction algorithm based on obstacle perception to construct a safe priority reachable area, and input the current position x of the captured drone. i、The location of the target escaping drone x e As well as the obstacle point cloud information Ω_obs in the environment, the hyperplane parameter set of the current safe priority reachable area for capturing the drone is obtained

[0109] 2-3) Hyperplanes for processing safe and accessible areas

[0110] right Each hyperplane H in is processed as follows:

[0111] Calculate the intersection line: Calculate the current hyperplane H and The intersection lines of other hyperplanes in are stored in lines;

[0112] Calculate intersection points and filter: Calculate the intersection points of each intersection line and only filter those in the safe and accessible area. The intersection points are added to the points collection;

[0113] Determine whether it constitutes a valid plane:

[0114] If the number of elements in the points set is less than 2, it means that these intersection points cannot form a valid plane. In this case, the hyperplane H is removed from It is removed from the convex polyhedron because it does not contribute to the formation of the final convex polyhedron;

[0115] At the same time, if the hyperplane H is the current capture drone position x i The target escape drone position x e The shared boundary hyperplane is set to False, indicating that there is no shared boundary between the current capture drone and the target;

[0116] If the number of elements in the points set is sufficient, a valid plane can be formed, and the points set is added to vertices_set;

[0117] 2-4) Determine the target point for the roundup

[0118] Determine the current capture target point of the drone based on the value of the shareplane flag:

[0119] If shareplane is True, it means that the current capture drone and the target have a shared boundary, and the centroid of the shared boundary is used as the capture target point, that is, T i is the centroid of the shared boundary;

[0120] If shareplane is False, it means that the current capture drone and the target do not share a boundary. eThe nearest point is taken as the target point, that is, T i is a convex polyhedron with x e nearest point;

[0121] 3) After the above steps, the corresponding capture target point is determined for each capture drone, and finally the capture target point set T is obtained. i as the output of the algorithm.

[0122] The multi-UAV cooperative capture method in an obstacle environment of the present invention has the following advantages:

[0123] (1) The three-dimensional decision space representation method based on the Voronoi diagram of the present invention can extract a convex bounded decision area (called the safety priority reachable area in the method of the present invention) from a typical non-convex area such as the three-dimensional obstacle space, thereby achieving effective representation of the decision space and laying the foundation for the generation of subsequent roundup strategies.

[0124] (2) The distributed capture decision method based on area minimization of the present invention can effectively protect the escaping drone by reducing the area of ​​the safe priority reachable area of ​​the escaping drone on the basis of the safe priority reachable area, and the present invention proves the theoretical completeness of this method. In addition, the present invention further realizes the lightweight of the algorithm by effectively sharing the computing load through the use of a distributed computing method. BRIEF DESCRIPTION OF THE DRAWINGS

[0125] Figure 1 It is a schematic diagram of the preferentially accessible area of ​​the present invention.

[0126] Figure 2 It is a schematic diagram of the safety priority accessible area of ​​the present invention.

[0127] Figure 3 It is a schematic diagram of constructing a safety priority reachable area in the simulation environment of the present invention. DETAILED DESCRIPTION

[0128] In order to better understand the purpose, structure and function of the present invention, the following is a further detailed description of a multi-UAV collaborative capture method in an obstacle environment of the present invention in conjunction with the accompanying drawings.

[0129] The present invention provides a method for cooperatively capturing multiple drones in an obstacle environment, which obtains spatial coordinate information of multiple drones including a defensive drone and an offensive drone based on relevant position capture equipment and algorithms.

[0130] Step 1: Three-dimensional decision space representation based on Voronoi diagram; The present invention is aimed at bounded convex regions containing obstacles. Such a three-dimensional obstacle space is a typical non-convex environment. It is very challenging to directly model and solve the control strategy. In order to meet this challenge, the present invention realizes a simplified representation of the three-dimensional obstacle decision space by constructing a safe priority accessible area. The method mainly includes the following steps:

[0131] Step 1.1: Construct a priority reachable domain based on the Voronoi diagram: Depending on the location of the capture drone cluster and the location of the escape drone, the Voronoi partition method is used in the three-dimensional scene to determine the priority reachable area of ​​the capture drone in the three-dimensional space.

[0132] Although the movement strategy of the escapee is inherently dynamic and unpredictable, it is still possible to capture the escapee in a limited time by effectively reducing the escapee's Priority Reachable Region (PRR). The set of all locations that can be reached faster than any other roundup.

[0133] Voronoi Diagram, also known as Thiessen polygon or Dirichlet diagram, is a geometric structure based on point set partitioning space. Its core idea is to divide a plane or three-dimensional space into multiple regions, and the distance from all points in each region to the seed point (generating point) of the region is smaller than the distance to other seed points. Specifically, given a set of seed points P = {p1, p2, ..., p n}, the Voronoi diagram divides the space into n convex polygonal regions, called Voronoi regions, or Dirichlet regions. Each Voronoi region V(p i ) consists of all the distances from the seed point p i The nearest point is formed, satisfying the following conditions:

[0134]

[0135] Among them, Q is the task scene space, x is any point in the space, and p i is the seed point, and the point in the Voronoi region to the seed point p i The distance to the seed point is smaller than the distance to other seed points.

[0136] In a two-dimensional plane, each Voronoi region in a Voronoi diagram is a convex polygon, while in three-dimensional space, the region is composed of polyhedrons. By dividing the space in this way, the Voronoi diagram can clearly show the relative position relationship between different seed points and their influence range. The Voronoi edge in the Voronoi diagram is the shared boundary between adjacent seed points, which is defined as the set of all points with equal distances to two adjacent seed points. In the three scenes, the shared Voronoi boundary is represented as the perpendicular bisector of the lines connecting the two seed points. Mathematically, a point x on the shared Voronoi boundary is g The following conditions are met:

[0137]

[0138] The construction of Voronoi diagram partition is relatively simple and computationally efficient, so it is very useful in many practical applications. In the multi-UAV cooperative capture problem described in the present invention, the application of Voronoi diagram helps to effectively divide the reachable area in the environment according to the locations of the capturers and escapees.

[0139] In an ideal environment without obstacles, for an agent system that follows a single integrator dynamics model and has equal maximum velocity, the escaper's priority reachable area can be represented by Voronoi partitioning. Specifically, the escaper's priority reachable area and the capturer's priority reachable area are defined as:

[0140]

[0141] in, and They represent the priority accessible areas of the capturer and escaper in an obstacle-free environment. In this environment, each agent (whether a capturer or an escaper) can determine its accessible area in space by calculating the Voronoi diagram, thereby optimizing its path planning and control strategy.

[0142] On the basis of clarifying that Voronoi partitioning can be used to construct a priority reachable domain, the present invention further realizes parameterized expression of the priority reachable domain by introducing the concept of hyperplane.

[0143] Hyperplane is an important geometric concept in high-dimensional space. It is a plane with one dimension lower than the space. Specifically, in d-dimensional space, the hyperplane is a subset of (d-1) dimensions. It can be used to divide the entire space into two half-spaces and is a very effective geometric separation tool. The basic feature of a hyperplane is that it divides any point in space into two categories: one on one side of the hyperplane and the other on the other side. In two-dimensional space, a hyperplane appears as a straight line; in three-dimensional space, a hyperplane is a two-dimensional plane.

[0144] A hyperplane in a three-dimensional environment can be defined by a normal vector and the coordinates of a point on the hyperplane. Suppose there is a known point P0 (x0, y0, z0) on the hyperplane and a normal vector Then the equation of the hyperplane can be expressed as:

[0145]

[0146] in, is an arbitrary point in space. The geometric object represented by this equation divides space into two half-spaces: one region satisfying n1(x-x0)+n2(y-y0)+n3(z-z0)>0, and the other region satisfying n1(x-x0)+n2(y-y0)+n3(z-z0)<0.

[0147] According to the hyperplane separation rule, the Voronoi partition of each trapper can be represented by a set of hyperplanes. Figure 1 As shown in Figure 1, the combination of the internal subregions of multiple hyperplanes can form a convex region, which represents the preferred reachable area of ​​the captor or escaper. Through this geometric construction, we can effectively describe and calculate the reachable area of ​​each agent in the environment.

[0148] Therefore, the definition of PRR can be transformed from formula (3) into the following form:

[0149]

[0150] Here, X represents the vector from the origin to any point x in Q, and l1 and l2 are used to refer to different agents including escapers and capturers. and are the hyperplane parameters defined by the position difference between agents l1 and l2, such that: n ei and p ei is the hyperplane parameter defined by the position difference between agents e and i, Represents the normal vector of the corresponding hyperplane and the plane position calculated by the midpoint.

[0151] In order to accurately describe the preferred reachable area of ​​each agent, the hyperplane normal vector and corresponding plane parameters associated with each pair of agents must be calculated. is determined by the relative position difference between the capturer or escaper, and the position parameter of the plane It is calculated by the midpoint between the two, ensuring that the hyperplane separates the regions according to the definition of Voronoi partition. The specific calculation formula is as follows:

[0152]

[0153] In the above formula, and denote the position vectors of agents l1 and l2 respectively, and is the relative position vector between them. Plane position parameters is determined by the midpoint of these two locations, ensuring that each hyperplane accurately separates the space.

[0154] Step 1.2: Construction of a safe priority reachable domain based on obstacle perception: Based on the priority reachable area, the safe distance between the captured drones and the minimum distance between the captured drones and obstacles are considered, and a safe priority reachable domain is constructed to ensure that these constraints are always met during the entire capture process.

[0155] Different from the existing methods that mainly focus on the assumption of obstacle-free scenarios, this paper expands the application scenarios to consider the actual environment containing obstacles. Based on the theoretical framework of priority safe accessible areas, we further propose a safe priority accessible area model based on obstacle perception to deal with the interaction between drones and obstacles in obstacle environments.

[0156] The separating hyperplane in formula (5) can effectively avoid collisions between multiple quadcopters, so we only need to further avoid collisions between drones and obstacles in the capture phase to ensure flight safety. To achieve real-time obstacle detection and avoidance, we construct a convex polyhedron to represent the safe flight area under the current position of the drone. The convex polyhedron is composed of multiple separating hyperplanes. Each hyperplane separates the drone from obstacles to form a safe and obstacle-free flight space. The mathematical representation of this space is as follows:

[0157]

[0158] in, Indicates the safe reachable area for drones. Defines the composition The set of parameters of the hyperplane. It should be noted that the number of hyperplanes is not fixed, but changes dynamically according to the geometry of the obstacles and the relative position of the drone. Unlike some studies that assume that obstacles are strictly convex regions with known vertices, we directly calculate the parameters of the separating hyperplane based on the point cloud Ω_obs constructed by the visual perception system of the quadrotor without making strict geometric assumptions on the obstacles.

[0159] In order to ensure the real-time decision-making, we use the KD-Tree data structure to accelerate the retrieval process of obstacle points. In this way, we can quickly obtain the obstacle point o_near closest to the current quadrotor position in the point cloud. Separating hyperplane parameters The calculation formula is as follows:

[0160]

[0161] Among them, r s is the safety radius, which is used to take into account the physical size of the drone and ensure the obstacle avoidance effect during flight by shifting the hyperplane away from the obstacle by a certain distance. Figure 2 As shown, by combining formula (5) and formula (7), the final representation of the safety priority reachable area can be defined as follows:

[0162]

[0163] This representation method ensures that the quadrotor can effectively avoid obstacles in complex environments, and by dynamically adjusting the number and positions of hyperplanes, it achieves adaptability to different obstacle forms, thus providing a solid theoretical foundation and computational support for real-time obstacle avoidance and flight decisions.

[0164] Step 1.3: Area iteration and update: Based on real-time scene changes (such as the relative position changes between the capture drone cluster and the escape drone), continuously adjust the safe priority accessible area to lay the foundation for high-frequency real-time re-decision-making.

[0165] When the relative positions of the capture drone cluster and the escape drone in the environment change, their safe priority reachable areas also change all the time. Therefore, it is necessary to recalculate the safe priority reachable areas based on the real-time changing position information. The specific calculation process is shown in Algorithm 1:

[0166] 1) Input information: current location x of the capture drone i , i∈Γ, the current position of the target escaping drone x e , obstacle point cloud information Ω_obs in the environment

[0167] 2) Initialize the remaining obstacle point cloud information Ω_remain=Ω_obs. Assign the obstacle point cloud information Ω_obs in the environment to the remaining obstacle point cloud information Ω_remain as the basis for subsequent processing.

[0168] 3) Constructing separation hyperplane based on obstacle point cloud

[0169] Enter a loop, and as long as the number of points in the remaining obstacle point cloud information is greater than a preset threshold, perform the following operations:

[0170] 3-1) Find the nearest obstacle point: Use the KD tree algorithm to find the nearest obstacle point to the current capture drone position x i The nearest obstacle point o_near.

[0171] 3-2) Construct obstacle separation hyperplane: calculate parameters Use these parameters to construct an obstacle separation hyperplane and store the relevant information of the hyperplane in the safe priority reachable area middle.

[0172] 3-3) Update the remaining obstacle point cloud information: remove the points outside the just constructed hyperplane from the remaining obstacle point cloud information to prepare for the next cycle.

[0173] 4) Dealing with other drones

[0174] Through a loop, all the capture drones (except the current i-th capture drone) are traversed, that is, j ranges from 0 to N and j is not equal to i:

[0175] 4-1) Set l1 as the number of the current captured drone, and l2 as the number of the currently traversed captured drone.

[0176] 4-2) Calculate parameters based on specific formulas <n ij , P ij > and store these parameters in the safe priority reachable area middle.

[0177] 5) Dealing with target escaping drones

[0178] Set l1 to the number of the current captured drone and l2 to the number of the target escaping drone. Calculate the parameters according to a specific formula <n ie , p ie > and store these parameters in the safe priority reachable area middle.

[0179] 6) After the above steps, the safe priority accessible area is finally obtained as the output of the algorithm.

[0180] Steps 1.2 to 1.3 are completed within milliseconds, successfully building a safe priority accessible area centered on each captured drone and escaping drone, and then further distributed decision-making is performed through step 2.

[0181] Step 2: Construct a distributed capture decision based on area minimization; based on the safe priority reachable area constructed by the three-dimensional decision space representation module based on the Voronoi diagram, a collaborative capture strategy based on area minimization is adopted to compress the movement area of ​​the escaping drone and finally achieve effective capture. This method mainly includes the following calculation processes:

[0182] Step 2.1: Construct a collaborative capture strategy based on area minimization: Based on the safe priority reachable domain constructed in step 1.2, the escaping drone is effectively captured by reducing the safe priority reachable area of ​​the escaping drone.

[0183] In step 1, the safe priority reachable area of ​​the escaping drone is defined as the set of spatial points that the escaping drone can reach before any of its surrounding drones. Therefore, the area of ​​the escapee's safe priority reachable area is is defined as:

[0184]

[0185] in is the safe priority accessible area of ​​the escapee constructed in formula (9). According to the definition of formula (10), The derivative of is:

[0186]

[0187] From the above formula, we can see that The movement strategy of the escaping drone and the movement strategy for capturing drones The movement strategy of the escaping drone is unknown to the capturing drone, but the capturing drone can adjust its movement strategy to gradually reduce the safe priority reachable area of ​​the evader over time. Finally, the capture is achieved. For further analysis, in formula (11), the capture of the drone The influence of can be decoupled as This decoupling allows the hunter to adopt an "area-minimizing" movement strategy that is consistent with The gradient descent direction is consistent, so as to achieve a rapid reduction in the area of ​​the escapee's safe priority reachable area. The movement strategy of the capturer is given by the following formula:

[0188]

[0189] The above formula describes the capture movement strategy of the drone. Next, the present invention will derive the specific calculation method of the strategy. Using the Leibniz integral rule, can be simplified to:

[0190]

[0191] Among them, N s represents the set of capturing drones i that share a boundary with the escaping drone e. i represents the shared boundary between the two. The area of ​​the shared boundary can be expressed as The centroid of the shared boundary is defined as

[0192] By comparing equation (11) and equation (13), we can conclude that

[0193]

[0194] Therefore, the area of ​​the decoupled safe priority reachable area of ​​the escaping drone is affected by the capturing drone as follows:

[0195]

[0196] Substituting equation (15) into equation (12), the final round-up movement strategy is simplified to:

[0197]

[0198] From a physical point of view, the "area minimization" collaborative capture strategy proposed by the present invention guides the capturer to the center of mass of the shared boundary between the capturer and the escapee. The core of this strategy is to gradually reduce the area of ​​the escapee's safe priority reachable area, and finally achieve the capture of the escapee.

[0199] Step 2.2: Prove the effectiveness of the round-up strategy: Prove the theoretical completeness of the collaborative round-up strategy based on area minimization proposed in step 2.1 through theoretical analysis;

[0200] Movement strategies in roundup drones aim to reduce the area of ​​escapees To prove that the strategy can guarantee capture, in the scenario of a single capture drone, we prove the safe reach area of ​​the escaping drone The distance between the capturing drone and the escaping drone is always strictly non-increasing, and the distance between the capturing drone and the escaping drone is also always strictly non-increasing. When extended to the scenario of multiple capturing drone swarms, these dynamic boundary conditions will lead to capture in a finite time.

[0201] Lemma 2: Consider a mission scenario where a single capture drone and a single escape drone move in region Q. According to the proposed capture strategy (16), for any allowed escape strategy, Furthermore, if and only if the escaper adopts the following movement strategy,

[0202]

[0203] Among them, C s Is a tracking drone x i and escape dronex e The centroid of the shared boundary of the safety-first reachable region.

[0204] Proof: For the case of a single pursuer and a single escaper, Simplified to:

[0205]

[0206] Substituting the area minimization-based round-up strategy (16) obtained in step 2.1 into the above formula, we obtain:

[0207]

[0208] Since the capture drone and the escape drone move at the same speed, the shared boundary between them is represented by the center of mass on the perpendicular bisector between them. is located on this shared boundary, so we have And in step 2.1, without loss of generality, the present invention assumes that v i =v e =1, that is Substituting the above conditions, we can see that for a single tracker, And when When , the control strategy of the escaper is:

[0209]

[0210] So far, Lemma 2 has been proved that the movement strategy of the hunter can ensure that the area of ​​the escapee's safe reachable area is strictly non-increasing, and the only escapee strategy that can keep the area of ​​the escapee's safe reachable area unchanged is to move in the opposite direction of the centroid of the shared boundary between the two.

[0211] The square of the distance between the escaper and the pursuer is defined as:

[0212] D=||x i -x e || 2 =(x i -x e ) T (x i -x e ) (twenty one)

[0213] Lemma 3: When the hunter adopts the movement strategy (16), the rate of change of the square of the distance satisfy In addition, if and only if the escaper adopts the control strategy shown in formula (20),

[0214] Proof: According to formula (21), the definition of D is

[0215]

[0216] Substituting the movement strategy (16) of the hunter into the above formula, we get:

[0217]

[0218] Also due to Located on the perpendicular bisector of the capture drone and the escape drone, so The above formula can be further derived as:

[0219]

[0220] Since the capture drone described in the present invention does not have a speed advantage, in step 2.1, without loss of generality, the present invention assumes that v i =v e =1, that is Therefore, there is and The motion strategy of the escaping drone is shown in formula (20).

[0221] So far, Lemma 3 has been proved. The movement strategy of the hunter can ensure that the distance between him and the escapee is strictly non-increasing. For the escapee, the only escape strategy that can keep the distance unchanged is to move in the opposite direction of the centroid of the shared boundary between the two.

[0222] According to Lemma 2 and Lemma 3, the present invention concludes that: in the scenario of a single capture drone, when the capture drone adopts the motion strategy shown in formula (16), the area of ​​the escapee's safe priority reachable area and the distance of the escapee from the capture drone are strictly non-increasing, and the rate of change of the two is zero only when the escapee moves in the opposite direction of the shared boundary centroid of the two, and in other cases both are decreasing. Therefore, returning to the multi-drone collaborative capture scenario described in the present invention, when the capture drone adopts the capture strategy described in formula (16), no matter what motion strategy the escapee adopts, it cannot guarantee that the condition that the distance from the escapee to each adjacent capture drone is not increasing is met at the same time. Therefore, the distance between the escapee and the capture drone cluster will always be in a decreasing state, and the capture drone cluster will eventually achieve final capture within a limited time.

[0223] Step 2.3: Distributed collaborative roundup decision framework: lightweight algorithm by allocating computational load.

[0224] In the multi-UAV collaborative capture decision method described in the present invention, the capture drone relies on environmental information, the positions of other capture drones in the cluster, and the position of the target escape drone as input information to complete the capture decision process. The present invention adopts a distributed capture decision framework. After each capture drone obtains the input information, it only constructs a safe priority reachable area centered on itself, and on this basis calculates the shared boundary centroid with the escape drone as the capture target point. In addition, considering that there may be a situation in the process of multi-UAV collaborative capture that there is no shared boundary between the current capture drone and the escape drone's safe priority reachable domain, at this time, the current capture drone's safe priority reachable domain degenerates into a safe reachable area. By adopting a greedy method, the closest point to the target escape drone is calculated in such a convex space as the target point, so as to get as close to the escape drone as possible.

[0225] The distributed collaborative roundup decision algorithm framework of the present invention is shown in Algorithm 2.

[0226] 1) Initialization

[0227] For each captured drone, initialize the following variables:

[0228] A set of hyperplane parameters used to store safe-priority reachable regions.

[0229] vertices_set: A set of vertices for storing each face of the convex polyhedron formed by the safe priority reachable area.

[0230] shareplane: A flag used to indicate whether the current capture drone and the target have a shared boundary.

[0231] 2) Traverse each captured drone

[0232] Through a loop, perform the following operations on each drone in the swarm:

[0233] 2-1) Reset the shared boundary flag: The shareplane flag is initially set to True, indicating that it is assumed that the current capture drone and the target have a shared boundary.

[0234] 2-2) Constructing a safe priority reachable area: Call the SPRRConstruction algorithm based on obstacle perception to construct a safe priority reachable area, and input the current position x of the captured drone. i 、The location of the target escaping drone x e As well as the obstacle point cloud information Ω_obs in the environment, the hyperplane parameter set of the current safe priority reachable area for capturing the drone is obtained

[0235] 2-3) Hyperplanes for processing safe and accessible areas

[0236] right Each hyperplane H in is processed as follows:

[0237] Calculate the intersection line: Calculate the current hyperplane H and The intersection lines of other hyperplanes in are stored in lines.

[0238] Calculate intersection points and filter: Calculate the intersection points of each intersection line and only filter those in the safe and accessible area. The intersection points are added to the points collection.

[0239] Determine whether it constitutes a valid plane:

[0240] If the number of elements in the points set is less than 2, it means that these intersection points cannot form a valid plane. In this case, the hyperplane H is removed from , since it does not contribute to the formation of the final convex polyhedron.

[0241] At the same time, if the hyperplane H is the current capture drone position x i The target escape drone position x e The shared boundary hyperplane is set to False, indicating that there is no shared boundary between the current capture drone and the target.

[0242] If the number of elements in the points set is sufficient (i.e., it can form a valid plane), add the points set to vertices_set.

[0243] 2-4) Determine the target point for the roundup

[0244] Determine the current capture target point of the drone based on the value of the shareplane flag:

[0245] If shareplane is True, it means that the current capture drone and the target have a shared boundary, and the centroid of the shared boundary is used as the capture target point, that is, T i is the centroid of the shared boundary.

[0246] If shareplane is False, it means that the current capture drone and the target do not share a boundary. e The nearest point is taken as the target point, that is, T i is a convex polyhedron with x e The nearest point.

[0247] 3) After the above steps, the corresponding capture target point is determined for each capture drone, and finally the capture target point set T is obtained. i as the output of the algorithm.

[0248] Compared with the centralized capture decision method that constructs all the safe priority reachable areas on a single drone and calculates the strategy before publishing it to other drones for execution, the distributed decision framework proposed in the present invention shares the computing load and reduces the decision time. Finally, the safe priority reachable area constructed by the method of the present invention in the simulation environment is as shown in the attached figure. Figure 3 As shown, the area formed by the blue squares represents the obstacle area Q o The red spherical quadrotor represents the pursuer, the blue spherical quadrotor represents the escaper, and the colored convex areas represent the safe priority reachable region (SPRR) of each quadrotor. and

[0249] Steps 2.1 to 2.3 are completed in polynomial time based on the three-dimensional decision space representation based on the Voronoi diagram. The entire decision process can meet the requirement of a decision frequency of 20 Hz, which meets the real-time requirements of the method deployed on the UAV.

[0250] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. A method for cooperatively capturing multiple drones in an obstacle environment, which obtains spatial coordinate information of multiple drones including a defensive drone and an offensive drone according to relevant position capture equipment and algorithms, and is characterized in that: The steps include: Step 1: Three-dimensional decision space representation based on Voronoi diagram: A simplified representation of the three-dimensional obstacle decision space is achieved by constructing a safe priority accessible area; Step 2: Construct a distributed capture decision based on area minimization: Based on the safe priority reachable area constructed by the three-dimensional decision space representation module based on the Voronoi diagram, a collaborative capture strategy based on area minimization is adopted to compress the movement area of ​​the escaping drone and ultimately achieve effective capture.

2. The method for cooperative capture of multiple UAVs in an obstacle environment according to claim 1 is characterized in that: The step 1 comprises the following steps: Step 1.1: Construct a priority reachable domain based on the Voronoi diagram: Depending on the location of the capture drone cluster and the location of the escape drone, the Voronoi partition method is used in the three-dimensional scene to determine the priority reachable area of ​​the capture drone in the three-dimensional space. Step 1.2: Construction of safe priority reachable domain based on obstacle perception: Based on the priority reachable area, the safe distance between the captured drones and the minimum distance between the captured drones and obstacles are considered. By constructing a safe priority reachable domain, it is ensured that these constraints are always met during the entire capture process. Step 1.3: Area iteration and update: Based on real-time scenario changes, continuously adjust the safe priority accessible areas to lay the foundation for high-frequency real-time re-decision-making.

3. The method for cooperative capture of multiple UAVs in an obstacle environment according to claim 2 is characterized in that: The step 1.1 includes the following specific steps: The escapee's priority accessible area is defined as the escapee's priority accessible area in a given environment. The set of all locations that can be reached faster than any other roundup; Given a set of seed points P = {p1, p2, ..., p n }, the Voronoi diagram divides the space into n convex polygonal regions, called Voronoi regions, or Dirichlet regions, each Voronoi region V(p i ) consists of all the distances from the seed point p i The nearest point is formed, satisfying the following conditions: Among them, Q is the task scene space, x is any point in the space, and p i is the seed point, and the point in the Voronoi region to the seed point p i The distance to is smaller than the distance to other seed points; In the two-dimensional plane, each Voronoi region of the Voronoi diagram is a convex polygon, while in three-dimensional space, the region is composed of polyhedrons. The Voronoi edge in the Voronoi diagram is the shared boundary between adjacent seed points, which is defined as the set of all points with equal distances to two adjacent seed points. In the three-dimensional scene, the shared Voronoi boundary is represented as the perpendicular bisector of the line connecting the two seed points. Mathematically, a point x on the shared Voronoi boundary is g The following conditions are met: The priority accessible areas for escapers and the priority accessible areas for capturers are defined as: in, and They represent the priority accessible areas of the captors and escapees in an obstacle-free environment. In this environment, each agent can determine its accessible area in space by calculating the Voronoi diagram, thereby optimizing its path planning and control strategy. The parameterized expression of the priority reachable domain is realized by introducing the concept of hyperplane: The hyperplane in a three-dimensional environment can be defined by the normal vector and the coordinate representation of a point on the hyperplane. Suppose there is a known point P0 (x0, y0, z0) on the hyperplane and the normal vector Then the equation of the hyperplane is expressed as: in, is an arbitrary point in space. The geometric object represented by this equation divides space into two half-spaces: one region satisfying n1(x-x0)+n2(y-y0)+n3(z-z0)>0, and the other region satisfying n1(x-x0)+n2(y-y0)+n3(z-z0)<0; According to the hyperplane separation rule, the Voronoi partition of each capturer can be represented by a set of hyperplanes. The combination of the internal sub-regions of multiple hyperplanes can form a convex region, which represents the priority reachable area of ​​the capturer or escaper. Therefore, the definition of the priority reachable region PRR can be transformed from formula (3) into the following form: Here, X represents the vector from the origin to any point x in Q, and l1 and l2 are used to refer to different agents including escapers and capturers. and are the parameters of the hyperplane defined by the difference in positions between agents l1 and l2, Represents the normal vector of the corresponding hyperplane and the plane position calculated by the midpoint; Calculate the hyperplane normal vector and corresponding plane parameters associated with each pair of agents. The normal vector of the hyperplane is determined by the relative position difference between the capturer or escaper, and the position parameter of the plane It is calculated by the midpoint between the two, ensuring that the hyperplane separates the regions according to the definition of Voronoi partition. The specific calculation formula is as follows: In the above formula, and denote the position vectors of agents l1 and l2 respectively, and is the relative position vector between them, and the plane position parameter is determined by the midpoint of these two locations, ensuring that each hyperplane accurately separates the space.

4. The method for cooperative capture of multiple UAVs in an obstacle environment according to claim 2, characterized in that: The step 1.2 includes the following specific steps: The safe flight area under the current position of the drone is represented by constructing a convex polyhedron, which is composed of multiple separating hyperplanes. Each hyperplane separates the drone from obstacles to form a safe and obstacle-free flight space. The mathematical representation of this space is as follows: in, Indicates the safe reachable area for drones. Defines the composition The parameter set of the hyperplane; The KD-Tree data structure is used to speed up the retrieval process of obstacle points and separate the hyperplane parameters. The calculation formula is as follows: Among them, r s is the safety radius, which is used to consider the physical size of the drone and ensure the obstacle avoidance effect during flight by offsetting the hyperplane a certain distance away from the obstacle. By combining formula (5) and formula (7), the final safe priority reachable area is defined as follows:

5. The method for cooperative capture of multiple UAVs in an obstacle environment according to claim 2, characterized in that: The step 1.3 includes the following specific steps: Recalculate the safe priority accessible area based on the real-time changing location information, 1) Input information: current location x of the capture drone i ,i∈Γ, the current position of the target escaping drone x e , obstacle point cloud information Ω_obs in the environment; 2) Initialize the remaining obstacle point cloud information Ω_remain=Ω_obs. Assign the obstacle point cloud information Ω_obs in the environment to the remaining obstacle point cloud information Ω_remain; 3) Construct a separating hyperplane based on the obstacle point cloud; Enter a loop, and as long as the number of points in the remaining obstacle point cloud information is greater than a preset threshold, perform the following operations: 3-1) Find the nearest obstacle point: Use the KD tree algorithm to find the nearest obstacle point to the current capture drone position x i The nearest obstacle point o_near; 3-2) Construct obstacle separation hyperplane: calculate parameters Use these parameters to construct an obstacle separation hyperplane and store the relevant information of the hyperplane in the safe priority reachable area middle; 3-3) Update the remaining obstacle point cloud information: remove the points outside the just constructed hyperplane from the remaining obstacle point cloud information to prepare for the next cycle; 4) Dealing with other drones Through a loop to traverse all the captured drones, that is, j ranges from 0 to N and j is not equal to i: 4-1) Set l1 as the number of the current captured drone, and l2 as the number of the currently traversed captured drone; 4-2) Calculate parameters based on specific formulas <n ij ,p ij > and store these parameters in the safe priority reachable area middle; 5) Dealing with target escaping drones Set l1 to the number of the current captured drone, l2 to the number of the target escaping drone, and calculate the parameters according to a specific formula <n ie ,p ie > and store these parameters in the safe priority reachable area middle; 6) After the above steps, the safe priority accessible area is finally obtained as the output of the algorithm.

6. The method for cooperative capture of multiple UAVs in an obstacle environment according to claim 2, characterized in that: The step 2 comprises the following specific steps: Step 2.1: Construct a collaborative capture strategy based on area minimization: Based on the safe priority reachable domain constructed in step 1.2, the escaped drone is effectively captured by reducing the safe priority reachable area of ​​the escaped drone; Step 2.2: Prove the effectiveness of the round-up strategy: Prove the theoretical completeness of the collaborative round-up strategy based on area minimization proposed in step 2.1 through theoretical analysis; Step 2.3: Distributed collaborative roundup decision framework: lightweight algorithm by allocating computational load.

7. The method for cooperative capture of multiple UAVs in an obstacle environment according to claim 6 is characterized in that: The step 2.1 includes the following specific steps: The area of ​​the escapee's safe priority accessible area is defined as: in is the safe priority accessible area of ​​the escapee constructed in formula (9). According to the definition of formula (10), The derivative of is: In formula (11), the capture of drones The influence of can be decoupled as With this decoupling, the hunter adopts an "area minimization" movement strategy that is consistent with The gradient descent direction is consistent with that of , which can achieve a rapid reduction in the area of ​​the escapee's safe priority accessible area. The movement strategy of the capturer is given by the following formula: Derived the specific calculation method of the above strategy, using the Leibniz integration rule, can be simplified to: Among them, N s represents the set of capturing drones i that share a boundary with the escaping drone e, s i represents the shared boundary between the two. The area of ​​the shared boundary can be expressed as The centroid of the shared boundary is defined as By comparing equation (11) and equation (13), we can get: Therefore, the area of ​​the decoupled safe priority reachable area of ​​the escaping drone is affected by the capturing drone as follows: Substituting equation (15) into equation (12), the final round-up movement strategy is simplified to: The "area minimization" collaborative capture strategy guides the capturer to the centroid of the shared boundary between the capturer and the escapee. This strategy gradually reduces the area of ​​the escapee's safe priority reachable area, and eventually achieves the capture of the escapee.

8. The method for cooperative capture of multiple UAVs in an obstacle environment according to claim 6, characterized in that: The step 2.2 includes the following specific steps: Lemma 2: Consider a mission scenario where a single capture drone and a single escape drone move in region Q. According to the proposed capture strategy (16), for any allowed escape strategy, Furthermore, if and only if the escaper adopts the following movement strategy, Among them, C s Is a tracking drone x i and escape dronex e The centroid of the shared boundary of the safety-first reachable area. Proof: For the case of a single pursuer and a single escaper, Simplified to: Substituting the area minimization-based round-up strategy (16) obtained in step 2.1 into the above formula, we obtain: Since the capture drone and the escape drone move at the same speed, the shared boundary between them is represented by the center of mass on the perpendicular bisector between them. is located on this shared boundary, so we have And without loss of generality in step 2.1, assume that v i =v e =1, that is Substituting the above conditions, we can see that for a single tracker, And when When , the control strategy of the escaper is: So far, Lemma 2 has been proved. The movement strategy of the hunter can ensure that the area of ​​the escapee's safe reachable area is strictly non-increasing, and the only escapee strategy that can keep the area of ​​the escapee's safe reachable area unchanged is to move in the opposite direction of the centroid of the shared boundary between the two. The square of the distance between the escaper and the pursuer is defined as: D=‖x i -x e ‖ 2 =(x i -x e ) T (x i -x e ) (21) Lemma 3: When the hunter adopts the movement strategy (16), the rate of change of the square of the distance satisfy In addition, if and only if the escaper adopts the control strategy shown in formula (20), Proof: According to formula (21), the definition of D is Substituting the movement strategy (16) of the hunter into the above formula, we get: Also due to Located on the perpendicular bisector of the capture drone and the escape drone, so The above formula can be further deduced as: Without loss of generality in step 2.1, the present invention assumes that v i =v e =1, that is Therefore, there is and The motion strategy of the escaping drone is shown in formula (20); So far, Lemma 3 has been proved. The movement strategy of the hunter can ensure that the distance between him and the escaper is strictly non-increasing. For the escaper, the only escape strategy that can keep the distance unchanged is to move in the opposite direction of the centroid of the shared boundary between the two. According to Lemma 2 and Lemma 3, it is concluded that in the scenario of a single capture drone, when the capture drone adopts the motion strategy shown in formula (16), the area of ​​the escapee's safe priority reachable area and the distance of the escapee from the capture drone are strictly non-increasing, and the rate of change of the two is zero only when the escapee moves in the opposite direction of the shared boundary centroid of the two, and in other cases both are decreasing; therefore, returning to the multi-drone collaborative capture scenario described in the present invention, when the capture drone adopts the capture strategy described in formula (16), no matter what motion strategy the escapee adopts, it cannot guarantee that the condition that its distance from each adjacent capture drone is not increasing is met at the same time, so the distance between the escapee and the capture drone cluster will always be in a decreasing state, and eventually the capture drone cluster will be finally captured within a limited time.

9. The method for cooperative capture of multiple UAVs in an obstacle environment according to claim 6, characterized in that: The step 2.3 includes the following specific steps: 1) Initialization For each captured drone, initialize the following variables: A set of hyperplane parameters used to store safe priority reachable areas; vertices_set: used to store the vertex set of each face of the convex polyhedron formed by the safe priority reachable area; shareplane: A flag used to indicate whether the current capture drone and the target share a boundary; 2) Traverse each captured drone Through a loop, perform the following operations on each drone in the swarm: 2-1) Reset the shared boundary flag: Initially set the shareplane flag to True, indicating that it is assumed that the current capture drone and the target have a shared boundary; 2-2) Constructing a safe priority reachable area: Call the SPRRConstruction algorithm based on obstacle perception to construct a safe priority reachable area, and input the current position x of the captured drone. i 、The location of the target escaping drone x e As well as the obstacle point cloud information Ω-obs in the environment, the hyperplane parameter set of the current safe priority reachable area for capturing the drone is obtained 2-3) Hyperplanes for processing safe and accessible areas right Each hyperplane H in is processed as follows: Calculate the intersection line: Calculate the current hyperplane H and The intersection lines of other hyperplanes in are stored in lines; Calculate intersection points and filter: Calculate the intersection points of each intersection line and only filter those in the safe and accessible area. The intersection points are added to the points collection; Determine whether it constitutes a valid plane: If the number of elements in the points set is less than 2, it means that these intersection points cannot form a valid plane. In this case, the hyperplane H is removed from It is removed from the convex polyhedron because it does not contribute to the formation of the final convex polyhedron; At the same time, if the hyperplane H is the current capture drone position x i The target escape drone position x e The shared boundary hyperplane is set to False, indicating that there is no shared boundary between the current capture drone and the target; If the number of elements in the points set is sufficient, a valid plane can be formed, and the points set is added to vertices_set; 2-4) Determine the target point for the roundup Determine the current capture target point of the drone based on the value of the shareplane flag: If shareplane is True, it means that the current capture drone and the target have a shared boundary, and the centroid of the shared boundary is used as the capture target point, that is, T i is the centroid of the shared boundary; If shareplane is False, it means that the current capture drone and the target do not share a boundary. e The nearest point is taken as the target point, that is, T i is a convex polyhedron with x e nearest point; 3) After the above steps, the corresponding capture target point is determined for each capture drone, and finally the capture target point set T is obtained. i as the output of the algorithm.

Images