Method and device for multi-uav path planning of threat area based on mubcr-rrt algorithm

Abstract

The application provides a threat area multi-unmanned aerial vehicle path planning method and device based on an MUBCR-RRT algorithm, and the method comprises the following steps: dividing a flight environment into grid units, and constructing a threat environment model; calculating a threat value of each grid unit according to the threat environment model; generating path points of the multi-unmanned aerial vehicle according to the threat value of the grid unit; and adopting Dubins to optimize the path points of the multi-unmanned aerial vehicle. The threat area and the building area are effectively distinguished, and the space in which the unmanned aerial vehicle can fly is increased. The unnecessary path points are reduced, and the path quality is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial vehicle path planning, in particular to a threat area multi-unmanned aerial vehicle path planning method and device based on a MUBCR-RRT algorithm. BACKGROUND

[0002] Due to the advantages in flexibility, cost-effectiveness and casualties, unmanned aerial vehicles play an important role in commercial, agricultural, military and other fields. Compared with a single unmanned aerial vehicle, a swarm of unmanned aerial vehicles has the characteristics of high efficiency, strong robustness and good scalability, and can perform complex tasks that a single unmanned aerial vehicle cannot complete, including reconnaissance, surveillance, attack, etc. However, the flight task of a swarm of unmanned aerial vehicles is becoming increasingly complex, and it still faces great challenges to find a set of optimal flight trajectories that meet the flight constraints.

[0003] In military wars, cities have always been the focus of both sides, and urban warfare is an important part of affecting the victory of both sides. The urban environment is often more complex, including tall and dense buildings, and there are often enemy fire points set at the edges of the buildings, which will generate a larger threat area in the urban environment. This has a great impact on the completion of the penetration mission of the unmanned aerial vehicle. In the complex urban environment considering the threat area, planning a set of feasible and conflict-free flight paths is crucial for the completion of the penetration mission. At present, in the face of the complex urban environment considering the threat area, the urban building area and the threat area are mixed together, and both of them are directly constructed as non-passable areas, it is difficult to distinguish between those areas that can pass although there is a threat and those areas that are completely prohibited from passing, and it is still a relatively difficult problem to plan the path of the swarm of unmanned aerial vehicles in the complex urban environment considering the threat area.

[0004] In summary, the existing technology has the following problems: the path planning of the swarm of unmanned aerial vehicles cannot distinguish the threat of the flight area, and the path planning is not perfect. SUMMARY

[0005] The purpose of the present application is to solve the problem of how to distinguish the threat of the flight path of the swarm of unmanned aerial vehicles to optimize the path planning.

[0006] To this end, in one aspect, the embodiments of the present application provide a threat area multi-unmanned aerial vehicle path planning method based on a MUBCR-RRT algorithm, which comprises the following steps:

[0007] dividing the flight environment into grid cells and constructing a threat environment model;

[0008] calculating the threat value of each grid cell according to the threat environment model;

[0009] dividing the environment space into safe passable space, passable space with threat and non-passable space according to the threat value of the threat unit.

[0010] generating multi-unmanned aerial vehicle flight path points in the safe passable space and the passable space with threats;

[0011] Dubins is adopted to optimize the multi-unmanned aerial vehicle flight path points.

[0012] In another aspect, the embodiment of the present application also provides a multi-unmanned aerial vehicle path planning device based on the MUBCR-RRT algorithm, comprising:

[0013] A construction unit is configured to divide a flight environment into grid units and construct a threat environment model.

[0014] A calculation unit is configured to calculate a threat value of each grid unit according to the threat environment model.

[0015] A division unit is configured to divide an environment space into a safe passable space, a passable space with threats and an impassable space according to the threat values of the threat units.

[0016] A planning unit is configured to generate multi-unmanned aerial vehicle flight path points in the safe passable space and the passable space with threats.

[0017] An optimization unit is configured to adopt Dubins to optimize the multi-unmanned aerial vehicle flight path points.

[0018] The above technical solution has the following beneficial effects: the present application can effectively distinguish a threat region from a building region by dividing the entire environment space into grid units. This distinction provides an important basis for subsequent multi-unmanned aerial vehicle flight path planning. To cope with the multi-unmanned aerial vehicle path planning challenge, the present application proposes a safety detection mechanism, extends the application range of the Rapidly-exploring Random Tree star algorithm to the multi-unmanned aerial vehicle path planning field, and proposes a node generation strategy, a candidate node evaluation strategy and a path reconnection mechanism to improve the RRT* algorithm. These measures aim to reduce the planning time and improve the path quality, thereby optimizing the path planning performance of the multi-unmanned aerial vehicle system.

[0019] The present application adopts the Dubins algorithm to smooth the trajectory, ensuring that the generated path not only meets the dynamic characteristics of the unmanned aerial vehicle, but also realizes smooth and continuous flight. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is a flowchart of the multi-unmanned aerial vehicle path planning method based on the MUBCR-RRT algorithm provided by the embodiment of the present application;

[0021] Figure 2A structure schematic diagram of a threat area multi-unmanned aerial vehicle path planning device based on an MUBCR-RRT algorithm is provided in the embodiment of the present application.

[0022] Figure 3 A first embodiment flow chart of a threat area multi-unmanned aerial vehicle path planning method based on an MUBCR-RRT algorithm is provided in the embodiment of the present application. DETAILED DESCRIPTION

[0023] The technical solutions in the embodiments of the present application will be clearly and completely described in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0024] In urban operations, unmanned aerial vehicles become representatives of new domain and new quality combat forces. How to perform multi-unmanned aerial vehicle path planning in a city environment considering threat areas is an important problem. Key elements and limiting factors of the problem are as follows:

[0025] Environment problem: An environment modeling method is needed to convert the real environment into a problem space in which unmanned aerial vehicles can perform path planning. A new environment modeling method is needed to distinguish threat areas from building areas in a city environment considering threat areas. The present application proposes a threat environment grid method to solve the problem.

[0026] Number of unmanned aerial vehicles: In multi-unmanned aerial vehicle path planning, multiple unmanned aerial vehicles take off from different starting points and fly to different ending points. In this process, collision avoidance between unmanned aerial vehicles is considered, and the blocking of buildings, the threat of firepower points, etc. are also considered.

[0027] In the embodiments of the present application, as shown in Figure 1 , a threat area multi-unmanned aerial vehicle path planning method based on an MUBCR-RRT algorithm is provided, which comprises the following steps:

[0028] S101: Dividing a flight environment into grid units and constructing a threat environment model;

[0029] The present application proposes a threat environment grid method, which can effectively distinguish threat areas from building areas and increase the space in which unmanned aerial vehicles can fly.

[0030] The path planning of UAV needs to model the environment first. Through the environment modeling, the actual environment is converted into a mathematical representation that can be processed by the computer, providing a running space for the path planning algorithm. The environment grid method is to grid the selected area, and divide the UAV flight area into a polygon area. The invention proposes a threat grid method, which first divides the urban environment of the target area into cubic units of equal size. The space in these grid units is abstracted as a point, and the position of each unit is located by the index of the three-dimensional matrix. Compared with the environment grid method, the threat grid method proposed in the invention proposes an innovative method to represent the threat level of each grid unit. The threat level of each grid unit is quantified by the element value in the three-dimensional matrix, and the size of the numerical value is proportional to the threat degree.

[0031] The threat matrix is shown in formula (1):

[0032]

[0033] t(x,y,z) represents the threat degree of grid unit (x,y,z). When the value of t(x,y,z) is equal to one, it means that this grid is not passable. In the invention, the range of the environment space is defined by the maximum value x max ,y max ,z max on the three coordinate axes, respectively, representing the limit boundary of the space in the x-axis, y-axis and z-axis directions. Any path of the UAV must be planned within the specified range.

[0034] The sources of threats mainly consider two types, as shown in Table 1. One is buildings, and the other is firepower points arranged in the city. The UAV cannot pass through the building area, but can pass through the firepower point coverage area. In the firepower point coverage area, the UAV will still be threatened. Considering these two types of threat sources, for grid unit P(x,y,z), its threat degree is defined as follows:

[0035]

[0036] wherein, represents the threat value of building i to grid (x,y,z), and Nb represents the number of buildings. represents the threat value of firepower point i to (x,y,z), and Nf represents the number of firepower points.

[0037] Table 1 Threat type

[0038]

[0039] For buildings, if a certain grid cell contains a building, the threat degree of the grid cell is set to one. Considering that the UAV needs to maintain a certain safety distance from obstacles during flight, a safety distance of one is set for the obstacle. If a grid contains an obstacle, in order to ensure safe flight of the UAV, the threat degrees of the eight adjacent grids, i.e. the upper, lower, left, right, front, rear, upper left and lower left grids, are also set to one.

[0040] In the threat grid method, the firing coverage of the fire point is simulated as a cone-like body, and the cone area is the fire coverage area. The closer the grid to the fire point, the greater the threat degree. The fire point has the following attributes as shown in Table 2.

[0041] Table 2: Fire point attributes

[0042]

[0043]

[0044] For a grid point P(x, y, z), the threat of the fire point is as follows:

[0045]

[0046] where P f represents the grid cell in which the fire point is located, v f represents the direction angle of the firing. If the grid (x, y, z) is both within the firing range angle of the fire point and within the firing distance of the fire point, it is within the coverage range of the fire point. The closer the grid point to the fire point, the greater the threat value; the farther the grid point to the fire point, the smaller the threat value.

[0047] In addition, when analyzing the coverage range of the fire point, the invention also considers the shielding of the building. Specifically, even if a grid is within the firing range, if it is shielded by a building, the fire point cannot cover the grid point.

[0048] S102: calculating the threat value of each grid cell according to the threat environment model;

[0049] The threat value of the grid cell is greater than or equal to 0 and less than or equal to 1. When the threat value of the grid cell is 1, it is impassable. The greater the threat value of the grid cell, the higher the probability of flight under fire attack.

[0050] S103: dividing the environment space into safe passable space, passable space with threat and impassable space according to the threat value of the threat cell;

[0051] S104: generating multi-UAV flight path points in the safe passable space and the passable space with threats;

[0052] First, initialize the vertex set, edge set, path set and flag. In the initialization of the vertex set, the initial nodes of all UAVs are put into the vertex set as different sets. Next, enter the loop to perform multi-UAV path planning. If all UAVs reach the end point, exit the loop. If this UAV has reached the target point, jump to the next UAV path search. The node_generation function is a new node generation strategy proposed in the application, and the node_generation function can only generate a new node and the x_nearest corresponding to the new node. The find_near function searches all points within a certain range in the extended tree of this UAV. The rewire function is responsible for checking the nodes within a certain range around the new node and evaluating whether the path can be optimized by taking these nodes as parent nodes.

[0053] After that, the probability_testing function proposed in the application is used to detect whether the UAV can connect the end point. If so, the flight path of this UAV is generated. Next, after finding all UAV paths, the path_reconnection function is used to reconnect the path points, reduce unnecessary path points, and improve path quality.

[0054] S105: using Dubins to optimize the multi-UAV flight path points.

[0055] The threat environment model is as follows:

[0056]

[0057] where t(x, y, z) represents the threat degree of the grid cell (x, y, z), x max represents the limit boundary value of the grid cell space in the x-axis direction, y max represents the limit boundary value of the grid cell space in the y-axis direction, and z max represents the limit boundary value of the grid cell space in the z-axis direction.

[0058] In the threat environment model:

[0059]

[0060] where, represents the threat value of building i to the grid cell (x, y, z), and Nb represents the number of buildings. Nf represents the number of firepower points;

[0061] wherein,

[0062]

[0063] wherein, P is a grid point, P f represents the position of the firepower point; v f represents the direction angle of the firing;

[0064] θ represents the included angle between the grid unit and the firepower point, that is, v f and the included angle, in degrees;

[0065] dis is the distance of .

[0066] According to the threat value of the grid unit, the path point of the multiple unmanned aerial vehicles is generated, and specifically includes:

[0067] The starting point of the unmanned aerial vehicle is added to the corresponding expansion tree by adopting the MUBCR-RRT* algorithm;

[0068] In the process of executing the multi-unmanned aerial vehicle path planning, in order to avoid the collision of the routes of the unmanned aerial vehicles, it is necessary to ensure that the flight trajectories of different unmanned aerial vehicles cannot occupy the same grid unit. Specifically, in the construction process of an unmanned aerial vehicle expansion tree, whenever a new edge is generated, it is necessary to judge whether the new edge and any existing edge of the expansion tree of the remaining unmanned aerial vehicles occupy the same grid, and if the same grid is occupied, the new edge cannot be added to the expansion tree.

[0069] The flight node is searched in the expansion tree until all the expansion trees reach the terminal point;

[0070] The starting point of the unmanned aerial vehicle, the flight node and the terminal point of the unmanned aerial vehicle are connected to obtain the flight path;

[0071] The Dubins curve is adopted to optimize the flight path.

[0072] The application also provides a threat area multi-unmanned aerial vehicle path planning device based on the MUBCR-RRT algorithm, as shown in the figure, which comprises: Figure 2 The construction unit 21 is used for dividing the flight environment into grid units and constructing a threat environment model;

[0073] The calculation unit 22 is used for calculating the threat value of each grid unit according to the threat environment model;

[0074]

[0075] ​The dividing unit 23 divides the environment space into safe passable space, passable space with threat and impassable space according to the threat value of the threat unit;

[0076] The planning unit 24 is used for generating multi-UAV flight path points in the safe passable space and the passable space with threat;

[0077] The optimization unit 25 is used for trajectory optimization of the multi-UAV flight path points by Dubins.

[0078] The threat environment model, specifically:

[0079]

[0080] Wherein, t(x, y, z) represents the threat degree of the grid unit (x, y, z), x max represents the limit boundary value of the grid unit space in the x-axis direction, y max represents the limit boundary value of the grid unit space in the y-axis direction, z max represents the limit boundary value of the grid unit space in the z-axis direction.

[0081] In the threat environment model:

[0082]

[0083] Wherein, represents the threat value of the building i to the grid unit (x, y, z), Nb represents the number of buildings, represents the threat value of the firepower point i to (x, y, z), Nf represents the number of firepower points;

[0084] Wherein,

[0085]

[0086] Wherein, P is a grid point, P f represents the position of the firepower point; v f represents the direction angle of the shooting;

[0087] θ represents the included angle between the grid unit and the firepower point, that is, the included angle between v f and , unit: degree;

[0088] dis is distance.

[0089] The threat value of the grid unit is greater than or equal to 0 and less than or equal to 1.

[0090] The planning unit, specifically includes;

[0091] adding a starting point of the UAV into a corresponding extended tree by using the MUBCR-RRT* algorithm;

[0092] finding a flight node in the extended tree until all the extended trees reach an ending point;

[0093] connecting the starting point of the UAV, the flight node and the ending point of the UAV to obtain a flight path;

[0094] optimizing the flight path by using a Dubins curve.

[0095] The application has the advantages that: the application can effectively distinguish a threat area from a building area by dividing the whole environment space into grid units. The distinction provides an important basis for subsequent multi-UAV flight path planning. To cope with the multi-UAV path planning challenge, the application proposes a safety detection mechanism, extends the application range of the Rapidly-exploring Random Tree star algorithm to the multi-UAV path planning field, and proposes a node generation strategy, a candidate node evaluation strategy and a path reconnection mechanism and other measures to improve the RRT* algorithm. The measures aim to reduce the planning time and improve the path quality, thereby optimizing the path planning performance of the multi-UAV system.

[0096] The application uses the Dubins algorithm to smooth the trajectory, ensures that the generated path not only meets the dynamic characteristics of the UAV, but also can realize smooth and continuous flight.

[0097] The above technical solutions of the embodiments of the application will be described in detail below in combination with specific application examples. Technical details not introduced in the implementation process can be referred to the related description in the foregoing.

[0098] Embodiment 1:

[0099] The application provides a threat area multi-unmanned aerial vehicle path planning method based on an MUBCR-RRT algorithm.

[0100] Multi-unmanned aerial vehicles successfully pass through a threat area in a city environment from a starting position to a target position to perform a task. A plurality of buildings and a plurality of enemy firepower points are distributed in the environment. The unmanned aerial vehicles cannot fly through the buildings, but can pass through the areas covered by the firepower points and the areas without threats. However, the unmanned aerial vehicles have a risk of being shot down when passing through the coverage of the firepower points. The closer the unmanned aerial vehicles are to the firepower points, the higher the risk of being shot down, and the higher the threat level. The application proposes an environment threat model for a city environment with a threat area, which can distinguish different threat levels of the city environment.

[0101] An environment threat generation model is provided.

[0102] The path planning of the unmanned aerial vehicles first needs to model the environment. Through environment modeling, the actual environment is converted into a mathematical representation that can be processed by a computer, providing a running space for the path planning algorithm. The environment grid method is to grid the selected area, and uniformly divide the unmanned aerial vehicle flight area into a polygon area. The application proposes an innovative threat grid method. The algorithm first divides the city environment of the target area into cubic units of equal size.

[0103] The space in these grid units is abstracted as a point, and the position of each unit is located through the index of a three-dimensional matrix. Compared with the environment grid method, the threat grid method proposed by the application proposes an innovative method to represent the threat level of each grid unit. The threat level of each grid unit is quantified by the element value in the three-dimensional matrix, and the size of the numerical value is proportional to the threat degree. The threat matrix is shown in formula (1).

[0104]

[0105] t(x, y, z) represents the threat degree of the grid cell (x, y, z), and when the value of t(x, y, z) is equal to one, it means that the grid is impassable. In the present application, the range of the environment space is defined by the maximum values x max ,y max ,z max on the three coordinate axes, which represent the extreme boundaries of the space in the x-axis, y-axis and z-axis directions, respectively. Any path of the UAV must be planned within the prescribed range.

[0106] The sources of threats mainly consider two types, as shown in Table 1, one is a building, and the other is a firepower point arranged in the city. The UAV cannot pass through the building area, but can pass through the firepower point coverage area, and in the firepower point coverage area, the UAV will still be threatened. Considering the two types of threat sources comprehensively, for the grid cell P(x, y, z), its threat degree is defined as follows:

[0107]

[0108] wherein, represents the threat value of the building i to the grid (x, y, z), and Nb represents the number of buildings. represents the threat value of the firepower point i to (x, y, z), and Nf represents the number of firepower points.

[0109] For buildings, as long as a grid cell contains a building, the threat degree of the grid cell is set to one. Considering that the UAV needs to maintain a certain safety distance from obstacles during flight, a safety distance of one is set for the obstacles. If a grid contains an obstacle, in order to ensure the safe flight of the UAV, the threat degrees of the eight grids (i.e. above, below, left, right, front, back, upper left, and lower left) adjacent to it are also set to one.

[0110] In the threat grid method, the firing coverage range of the firepower point is simulated as a cone-like body, and the cone area is the firepower coverage area. The closer the grid is to the firepower point, the greater the threat degree it receives. The firepower point has the following attributes as shown in Table 2.

[0111]

[0112] wherein P f represents the grid cell in which the firepower point is located, and v f represents the direction angle of the firing. If the grid (x, y, z) is both within the firing range angle of the firepower point and within the firing distance of the firepower point, it is within the coverage range of the firepower point. And the closer the grid point is to the firepower point, the greater the threat value it receives; the farther the grid point is from the firepower point, the smaller the threat value it receives.

[0113] In addition, the present application also considers the shelter of buildings when analyzing the coverage range of the fire point. Specifically, even if a grid is within the shooting range, the fire point cannot cover the grid if it is sheltered by a building.

[0114] Path representation:

[0115] Parametric expression of the UAV path is of great significance to the study of UAV path planning. For a UAV, its path can be expressed as:

[0116] path={P0,P1,P2,…P k ,…,P n ,P n+1}(6)

[0117] where P_0 represents the starting point coordinate, P_(n+1) represents the end point. P_k=(x_k,y_k,z_k) represents the intermediate path points passed by the UAV during flight. These path points are discrete in the present application.

[0118] The current path expression form has certain deficiencies. For example, the path constructed by directly connecting the path points may result in a flight trajectory that is not smooth, and this non-smoothness may limit the performance of the UAV. There are two methods that can be used to smooth the path. One method is to use B-Spline curves to smooth the broken line formed by connecting the path points. This method can make the broken line smooth. However, this smoothing curve method may not necessarily meet the constraints of the UAV's own direction angle and pitch angle, and may also change the path passed by the UAV, causing it to collide with obstacles. The other method is to use the Dubins algorithm to construct a three-dimensional path between two obstacle-free points. This method can meet the performance constraints of the UAV and also try not to change the flight trajectory of the UAV.

[0119] The second method is used to connect the barrier-free points of the UAV flight trajectory. The initial Dubins curve is proposed to find the shortest path connecting two points in a two-dimensional plane while satisfying the curvature constraint of the vehicle. This ensures that the vehicle maintains a stable state during travel and can smoothly traverse the path between the two points. In three-dimensional space, in addition to the heading angle constraint, the UAV is also subject to the pitch angle constraint. Currently, there are various methods for generating three-dimensional Dubins paths, such as adding intermediate arcs to the path or using nonlinear methods to generate curves. This study adopts an accurate algorithm proposed by Hota et al., which can generate a path that satisfies the curvature and angle constraints in three-dimensional space for the UAV. Given the starting point P_start = (x_0, y_0, z_0) and the ending point P_end = (x_1, y_1, z_1) as well as the minimum turning radius R of the UAV, the optimal flight path between the two points can be determined by providing the heading angle and pitch angle of the starting and ending points.

[0120] The heading angle of the UAV represents its turning ability in the horizontal plane, i.e., the lateral offset angle of the UAV relative to the forward direction, denoted by γ. In the assumption of Dubins curve, the UAV is limited to flying only in its forward direction, so the range of the heading angle is Correspondingly, the pitch angle of the UAV describes its attitude adjustment in the vertical plane, i.e., the inclination angle upward or downward relative to the horizontal plane, denoted by ψ. Similarly, the range of the pitch angle is

[0121] There is a continuous path point P k , P k+1 on the path, and the flight direction of the UAV at the path point P k is calculated as follows:

[0122]

[0123] In the above formula, V k represents the flight direction of the UAV at P k . In particular, the flight direction of the UAV at the ending point P n+1 is consistent with the flight direction of the previous path point.

[0124] Assuming V k = (v xk , v yk , v zk ), according to the three-dimensional dynamics model of the UAV, the vector relationship between the heading angle and the pitch angle is as follows:

[0125] v xk = ||V k ||cosγcosψ(8)

[0126] v yk =||V k ||cosγsinψ(9)

[0127] v zk =||V k ||sinγ(10)

[0128] Therefore, the calculation formula of the direction angle and the pitch angle can be obtained as follows:

[0129]

[0130] Multi-UAV path planning model:

[0131] Multi-UAV three-dimensional path planning refers to finding a flight path that meets the conditions for multiple UAVs in three-dimensional space by considering various constraint conditions and task requirements. Multi-UAV path planning is a constraint optimization problem (COP). Generally, the basic form of the constraint optimization problem is as follows:

[0132] minf(x)(13)

[0133]

[0134] Where x=(x1,x2,…x D )εΩ is the decision variable, and Ω is the decision space formed by the decision vector. f(x) is the objective function, which maps him to the target space π. g j is an inequality constraint, and h j is an equality constraint. U i and L i are the upper and lower boundaries of the decision variable x i . And q is the total number of constraints.

[0135] For the multi-UAV path planning problem in the urban environment with threat areas, the goal is to minimize the length of the UAV flight path and prefer to pass through the grid points with lower threat levels. The length of the Dubins curve is mainly composed of two parts: the first part is the distance of the turn, and the second part is the distance of the straight line. Generally, the turning radius of the rotor UAV is very small, and the length of the first part is much smaller than the length of the second part. Therefore, the Euclidean distance is used to replace the length of the Dubins curve, and the objective function can be expressed as:

[0136]

[0137]

[0138] where i is an index from 1 to Nu, Nu represents the number of UAVs. j is a discrete point index from 0 to N, representing the order number of the UAV path point. ij The threat value of path point P ij is defined as the average of the threat values of the series of grids that the UAV passes through in the process of flying from grid point P ij to P ij+1 . k represents an index from 1 to Np ij , Np ij represents the number of grids that the i-th UAV passes through in the process of flying from path point P ij to P ij+1 .

[0139] In addition to defining the objective function, it is also important to properly define the UAV flight constraints. In the process of performing flight tasks in low altitude, the UAV group is generally required to penetrate from low altitude, so an upper limit of height, called the maximum height limit, is set, denoted by z max . In addition, the height of the UAV needs to be greater than the ground height. Therefore, the height constraint of the UAV is as follows:

[0140] 0 ≤ z ijk ≤ z max (18)

[0141] Multi-UAV path planning generally seeks feasible paths in a specified area, so the coordinate range of the grid points passed through by the UAV satisfies the constraint:

[0142] 0 ≤ x ijk ≤ x max (19)

[0143] 0 ≤ y ijk ≤ y max (20)

[0144] The UAV needs to avoid buildings during flight, i.e., the threat value of the grid passed through by the UAV cannot be 1. For the i-th UAV, its planned path has two path points P ij and P ij+1 . In the process of flying from P ij to P ij+1 , a series of grids will be passed through. The anti-collision constraint of the UAV and the building is as follows:

[0145]

[0146] where Np ij represents the grid index passed through by the i-th UAV in the process of flying from path point P ij to P ij+1 , t(x ijk , y ijk , z ijkrepresents the threat value of the grid point (x ijk ,y ijk ,z ijk ). Nu represents the number of UAVs, N i represents the number of path points on the i-th UAV trajectory.

[0147] The UAV needs to meet its own performance constraints, and the model mainly considers the direction angle and the pitch angle of the UAV during flight, which should not exceed the defined maximum range, so the two constraints are defined as follows:

[0148] g1=γ-γ max (24)

[0149] g2=ψ-ψ max (25)

[0150] γ max represents the maximum direction angle, ψ max represents the maximum pitch angle. Finally, for the UAV formation flight, the collision avoidance problem between UAVs also needs to be considered in the path planning process. The UAV collision constraint is designed as follows:

[0151]

[0152] In the above formula, cum ij represents whether the i-th UAV collides with the rest of the UAVs in the process of flying from path point P ij to P ij+1 .

[0153] Based on the above discussion, the multi-UAV path planning model can be described as follows:

[0154] minf=f(28)

[0155]

[0156] 0≤x ijk ≤x max (30)

[0157] 0≤y ijk ≤y max (31)

[0158] 0≤z ijk ≤z max (32)

[0159] Obviously, it is a constrained optimization problem. If the UAV trajectory can meet the constraints of (29)-(32), a set of required paths can be generated.

[0160] The present application proposes a Multi-Unmanned Aerial Vehicle Goal-Biased Candidate Evaluation Strategy with Reconnection Rapidly-exploring Random Tree star (MUBCR-RRT*) based on the RRT* algorithm. The following introduces the overall framework of MUBCR-RRT*, and the pseudo code of the algorithm is shown in Algorithm 1 in Table 3.

[0161] First, the vertex set, edge set, path set and flag are initialized, see line 1-7, and when initializing the vertex set, the initial nodes of all unmanned aerial vehicles are put into the vertex set as different sets. Next, enter the loop, see line 10, and perform multi-unmanned aerial vehicle path planning. If all unmanned aerial vehicles reach the destination, exit the loop, see line 11-12. If this unmanned aerial vehicle has reached the target point, jump to the path search of the next unmanned aerial vehicle, see line 13-15. The node_generation function is a new node generation strategy proposed by the present application, see Algorithm 3 for details. It should be noted that the node_generation function can only generate a new node and the x_nearest corresponding to the new node x_new, see line 16. The find_near function searches for all points within a certain range of x_new in the extension tree of this unmanned aerial vehicle, see line 18. The rewire function is responsible for checking the nodes within a certain range of the new node x_new and evaluating whether the path can be optimized by taking x_new as the parent node of these nodes, see line 21.

[0162] After that, the probability_testing function proposed by the present application is used to detect whether the unmanned aerial vehicle can connect the terminal point, and if so, the flight path of this unmanned aerial vehicle is generated, see line 23-25, and the specific steps of the probability_testing function are shown in Algorithm 5. Next, after finding all unmanned aerial vehicle paths, the path_reconnection function is used to reconnect the path points, see line 30, to reduce unnecessary path points and improve path quality, see Algorithm 6 for details.

[0163] Table 3 Algorithm 1

[0164]

[0165]

[0166] Compared with the traditional RRT* algorithm, MUBCR-RRT* has made the following improvements:

[0167] First, the algorithm plans the paths of multiple drones simultaneously, that is, it generates multiple expansion trees at the same time. Figure 3 As shown in the figure, a safety detection mechanism is introduced during the node connection phase. This mechanism, taking into account obstacles and fire points, performs collision detection on newly generated path segments and existing expansion trees. This ensures that drones can effectively avoid each other during flight, ensuring flight safety. This mechanism is applied in the new node connection phase, the probabilistic detection mechanism, and the path reconnection strategy.

[0168] Second, a node generation strategy is proposed. Specifically, multiple sampling points are first generated, then offset using a sampling bias strategy. Multiple candidate points are then generated through expansion. Afterwards, a candidate point evaluation strategy is used to evaluate and rank the candidate points, selecting the optimal nodes that ensure safety and adding them to the expansion.

[0169] Third, drawing on the principles of the artificial potential field method, a sampling bias strategy is proposed. An attractive potential field is set for the end point to attract sampling points closer to the end point, aiming to improve the efficiency of path search. By guiding sampling points to offset to more promising areas, the convergence speed of path planning is accelerated.

[0170] Fourth, by drawing on and optimizing the fitness function of the A* algorithm, we proposed a candidate point evaluation strategy and designed a new fitness function. This strategy evaluates candidate nodes, taking into account the positions and paths of other drones. This strategy improves path quality and further accelerates algorithm convergence.

[0171] Fifth, a probabilistic check is introduced to determine whether the path between the endpoint and the nearest node in the current tree is safe. If the path satisfies the safety check, the algorithm directly connects the nearest node to the endpoint, skipping the process of generating additional nodes in between.

[0172] Sixth, a path reconnection mechanism is proposed. After obtaining a feasible path, the paths of multiple drones are reconnected. On the basis of ensuring safety, unnecessary nodes are reduced and the path quality is improved.

[0173] Seventh, for the final path, the Dubins curve is used to optimize the trajectory so that it meets the UAV's self-generated performance constraints.

[0174] Safety detection mechanism:

[0175] In the process of multi-UAV path planning, in order to avoid the collision of UAV routes, it is necessary to ensure that the flight trajectories of different UAVs cannot occupy the same grid cell. Specifically, in the process of constructing the expansion tree of a UAV, whenever a new edge is generated, it is necessary to determine whether the new edge and any existing edge of the expansion tree of the remaining UAVs occupies the same grid. If the same grid is occupied, the new edge cannot be added to the expansion tree.

[0176] To solve the above problems, the present application proposes a safety detection mechanism, the pseudo code of which is shown in Algorithm 2 in Table 4. First, the mechanism checks the passability of the two points to be connected (see lines 1-2), and the pass_free function is used to ensure that there is no obstacle between the two points and that it can pass. Next, the present application introduces a distance parameter check_dis, which is directly related to the distance between x1 and x2 (see line 4). Then, all nodes in different expansion trees that are within a distance of check_dis from x1 are found (see line 9), and for these nodes, the present application performs intersection detection (see line 10). Specifically, it is determined whether the lines between these nodes and their parent and child nodes occupy the same grid as the new line to be generated. Finally, it can be determined whether the two points can be safely connected.

[0177] Table 4 Algorithm 2

[0178]

[0179] Node generation strategy:

[0180] Although the RRT* algorithm effectively solves the path planning problem, its random sampling-based nature leads to slow convergence speed and the risk of falling into a local optimal solution. Therefore, the present application designs a node generation strategy to generate new nodes for the random expansion tree. The specific steps of this strategy are shown in Algorithm 3 in Table 5. First, in the sampling point generation stage of the algorithm, not only one sampling point is generated, but multiple sampling points are generated (see lines 3-8), and a sampling bias strategy is designed to bias these sampling points (see line 4), resulting in multiple biased points. Then, multiple candidate points are generated based on these biased points (see line 6), and the steer function represents the expansion of the expansion number from x_nearest to the sampling point x_rand in the direction of step to obtain a new candidate point x_new. For these candidate points, the present application designs a candidate point evaluation strategy to evaluate and rank these candidate points (see line 9), and finally selects the optimal candidate point that can pass the safety detection mechanism as the new node of the tree (see lines 11-13).

[0181] Table 5 Algorithm 3

[0182]

[0183] Sampling bias strategy:

[0184] The present application draws on the idea of APF algorithm, and constructs an attractive potential field for the end point. After the random sampling point is generated, it is guided by the potential field, so as to obtain a bias point. The artificial potential field function set by the present application is shown in formula (34), wherein ζ is the gravitational potential field coefficient.

[0185]

[0186] The present application combines the conical and quadratic potential fields. At a long distance, the conical potential field provides a fixed larger attractive force, so that the intelligent agent can quickly move towards the target. When close, the quadratic potential field provides an attractive force that decreases as the distance decreases, which helps the object to approach the target more accurately. Combining the two potential fields, a potential field can be created that provides sufficient attractive force when the object is far from the target, and provides more delicate control when the intelligent agent approaches the target, thereby avoiding the overshooting problem.

[0187] The gradient corresponding to the potential field is shown in formula (35).

[0188]

[0189] wherein, is the distance at which the two potential fields are divided, and when the gradients of the quadratic potential field and the conical potential field are equal. In the present application is set to three times the step size, and ζ is set to 1 / 3.

[0190] Next, according to the random sampling point and the end point position, the attractive force suffered by the random sampling point is calculated, and the calculation formula is as follows:

[0191]

[0192] Candidate point evaluation strategy:

[0193] The present application innovatively proposes a candidate point evaluation strategy, which aims to evaluate and sort a plurality of candidate tree nodes. The strategy uses an optimized heuristic function to evaluate the value of each candidate node. After the evaluation is completed, all candidate tree nodes will be sequentially arranged from low to high according to their fitness values.

[0194] In the A* algorithm, the definition of its heuristic function is shown in formula (37):

[0195] f(x)=h(x)+g(x)(36)

[0196] where f(x) represents the value information of the current node x, h(x) represents the estimated cost from the current node to the terminal point, and g(x) represents the actual cost from the starting point to the current node.

[0197] The present application proposes an improved heuristic function, which is defined as shown in formula (37), assuming that there are Nu unmanned aerial vehicles, and thus Nu extended trees, and for an extended tree, the heuristic function thereof is:

[0198]

[0199] where F(x new ) represents the fitness value of the candidate node x new . x parent represents the parent node of x new , and t(x parent , x new ) represents the threat value from x parent to x new , and the specific calculation formula is shown in formula (38). f1(x new ) represents the value of x new in the current tree, and the specific calculation formula is shown in formula (39). In order to comprehensively evaluate the advantages and disadvantages of the existing node, ρ(x new ) is introduced, which represents the value of x new for other trees, and the specific calculation formula is shown in formula (40). At the same time, the weight α is introduced to adjust the importance of f1(x new ) and ρ(x new ) in the node selection process.

[0200] t(x parent , x new ) = min(1, δt passnode / num passnode )(38)

[0201] In formula (38), δ represents a weight coefficient of the threat value of the grid, which is used to quantify the importance of the threat area. t passnode represents the sum of the threat values of each grid on the path from x parent to x new , and num passnode represents the number of grids.

[0202]

[0203] In formula (39), ρ(x new ) represents the cumulative fitness value of x_new in other trees. f2(x new) to quantify the fitness value of x_new in other trees, which is calculated as shown in equation (41).

[0204] f1(x new ) = g(x parent ) + βh(x new ) + dis(x parent , x new ) (40)

[0205] In equation (40), f1(x new ) is composed of three parts. The first part is the path length consumed from the start point to x parent , which is denoted by g(x parent ). The second part is the distance from the candidate node x new to the end point, which is calculated using the Euclidean distance and denoted by h(x new ). The third part is the distance from x parent to x new , which is also calculated using the Euclidean distance and denoted by dis(x parent , x new ). β is the balancer of this function, which coordinates the dependence on the historical path and the response to the proximity to the target.

[0206]

[0207] Equation (41) introduces the lastx new parameter, which stores the newly added nodes in different expansion trees in the last iteration. represents the newly added nodes in the jth expansion tree in the last iteration. Thus, f2(x new ) introduces the value of the current node x_new for other expansion trees by replacing x new with in f1(x parent ). In order to make f2(x new ) more sensitive to the distance between and x new , the first half of f2(x new ) is introduced, where r safe represents the safety distance of the UAV, represents the Euclidean distance between the current candidate node and the newly added nodes in the last iteration, and γ is the weight coefficient to balance the weights between the two, which is set to 1 in the present application.

[0208] The candidate point evaluation strategy pseudo code is shown in Algorithm 4 in Table 6. First, find_near function is used to search all the points within a certain range from x_new in the current expansion tree. Next, find a new parent node for x_new, see line 3, and the role of chooseparent function is to filter the nodes in x_near_list to determine whether there is a node that can be the parent node of x_new, so as to optimize the path length from the starting point to x_new. Then, calculate the fitness value of each candidate node, see line 4-6, and sort it from small to large, see line 9, and finally return the sorted list.

[0209] Table 6 Algorithm 4

[0210]

[0211]

[0212] Probability detection strategy:

[0213] The termination condition of the algorithm is improved, and a probability detection mechanism is introduced. Specifically, for each tree, the strategy detects whether the nearest node to the target point can pass the safety detection through probability evaluation. Once the detection is passed, the algorithm will directly connect the point to the target point, thereby significantly improving the convergence efficiency of the algorithm.

[0214] The specific steps are shown in Algorithm 5 in Table 7. First, probability judgment is performed, see line 1, and then the nearest node to the end point is found, see line 2. Next, it is judged whether the nearest node and the end point can pass the safety check, see line 3. If yes, the final path is found, see line 4, and the path is returned, see line 5.

[0215] Table 7 Algorithm 5

[0216]

[0217] Path reconnection mechanism:

[0218] The application proposes a path reconnection machine to reduce unnecessary nodes on the path and further improve the path quality. The specific operation is shown in Algorithm 6 in Table 8. All paths are traversed one by one, and the starting point of each path is checked to see if it can be connected to the node behind it through safety inspection, directly skipping some nodes in between. If a node that fails the safety inspection is encountered, the x_parent node will jump to the previous node of the node, and the above operation will be repeated, as shown in lines 4-15, to realize the local optimization of the path. This process will continue along the entire path until the evaluation and possible reconnection of all nodes are completed.

[0219] Table 8 Algorithm 6

[0220]

[0221] The application can effectively distinguish between threat areas and building areas by dividing the entire environment space into grid cells. This distinction provides an important basis for subsequent multi-UAV flight path planning. To address the multi-UAV path planning challenge, the application proposes a safety detection mechanism, extends the application range of the Rapidly-exploring Random Tree star algorithm to the multi-UAV path planning field, and proposes node generation strategies, candidate node evaluation strategies, and path reconnection mechanisms to improve the RRT* algorithm. These measures aim to reduce planning time and improve path quality, thereby optimizing the performance of multi-UAV system path planning.

[0222] The application uses the Dubins algorithm to smooth the trajectory, ensuring that the generated path not only meets the dynamic characteristics of the UAV, but also realizes smooth and continuous flight.

[0223] The above specific embodiments further detail the purpose, technical solutions, and beneficial effects of the application. It should be understood that the above description is only a specific embodiment of the application and does not limit the protection scope of the application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the application should be included in the protection scope of the application.

Claims

1. A multi-UAV path planning method in a threat area based on the MUBCR-RRT algorithm is characterized by: The method comprises the following steps: Divide the flight environment into grid cells and build a threat environment model; Calculating a threat value for each grid cell according to the threat environment model; According to the threat value of the threat unit, the environment space is divided into safe traversable space, threatening traversable space and impassable space; Generate multi-drone flight path points in safe and threatening traversable spaces; Use Dubins to optimize the trajectory of multiple UAV flight path points; The threat environment model is specifically: Among them, t(x,y,z) represents the threat level of the grid cell (x,y,z), x max Represents the limit boundary value of the grid cell space in the x-axis direction, y max Represents the limit boundary value of the grid cell space in the y-axis direction, z max Represents the limit boundary value of the grid unit space in the z-axis direction; In the threat environment model: in, represents the threat value of building i to grid unit (x, y, z), Nb represents the number of buildings, represents the threat value of firepower point i to (x, y, z), and Nf represents the number of firepower points; in, Among them, P is the grid point, P f Indicates the location of the firing point; v f Indicates the direction angle of the shooting; θ represents the angle between the grid unit and the firing point, that is, v f and The angle between , in degrees; dis is distance.

2. The method for multi-UAV path planning in a threat area based on the MUBCR-RRT algorithm according to claim 1 is characterized in that: The threat value of the grid cell is greater than or equal to 0 and less than or equal to 1.

3. The method for multi-UAV path planning in a threat area based on the MUBCR-RRT algorithm according to claim 1 is characterized in that: Generate multi-UAV path points based on the threat value of the grid cell, including: The MUBCR-RRT* algorithm is used to add the starting point of the UAV to the corresponding expansion tree; Search for flight nodes in the expansion tree until all expansion trees reach the end point; Connecting the starting point of the drone, the flight node, and the end point of the drone to obtain a flight path; The flight path is optimized using the Dubins curve.

4. A multi-UAV path planning device in a threat area based on the MUBCR-RRT algorithm, characterized in that: include: Construction unit, used to divide the flight environment into grid units and build a threat environment model; a calculation unit, configured to calculate a threat value of each grid cell according to the threat environment model; Divide the unit into units, and divide the environment space into safe traversable space, threatening traversable space and impassable space according to the threat value of the threat unit; A planning unit for generating multi-UAV flight path points in safe traversable spaces and threatening traversable spaces; Optimization unit, used to optimize the trajectory of multiple UAV flight path points using Dubins; The threat environment model is specifically: Among them, t(x,y,z) represents the threat level of the grid cell (x,y,z), x max Represents the limit boundary value of the grid cell space in the x-axis direction, y max Represents the limit boundary value of the grid cell space in the y-axis direction, z max Represents the limit boundary value of the grid unit space in the z-axis direction; In the threat environment model: in, represents the threat value of building i to grid unit (x, y, z), Nb represents the number of buildings, represents the threat value of firepower point i to (x, y, z), and Nf represents the number of firepower points; in, Among them, P is the grid point, P f Indicates the location of the firing point; v f Indicates the direction angle of the shooting; θ represents the angle between the grid unit and the firing point, that is, v f and The angle between , in degrees; dis is distance.

5. The device for multi-UAV path planning in a threat area based on the MUBCR-RRT algorithm according to claim 4 is characterized in that: The threat value of the grid cell is greater than or equal to 0 and less than or equal to 1.

6. The device for multi-UAV path planning in a threat area based on the MUBCR-RRT algorithm according to claim 4 is characterized in that: The planning unit specifically includes: Used to add the starting point of the drone to the corresponding expansion tree using the MUBCR-RRT* algorithm; Used to find flight nodes in the expansion tree until all expansion trees reach the end point; Connecting the starting point of the UAV, the flight node, and the end point of the UAV to obtain a flight path; Used to optimize flight paths using the Dubins curve.

Images