Heterogeneous robot cooperative three-dimensional inspection planning method under vertical structure facility

Abstract

The invention relates to the technical field of robots, in particular to a heterogeneous robot cooperative three-dimensional inspection planning method under a vertical structure facility. Obtaining point cloud data of a target facility of a vertical structure; according to the point cloud data, a three-dimensional convex hull enhancement algorithm is adopted to generate a target facility minimum convex polygon describing a vertical structure; determining a cruise starting point and a cruise ending point of the unmanned aerial vehicle based on the minimum convex polygon; performing path planning between the cruise starting point and the cruise ending point of the unmanned aerial vehicle by adopting an improved A star algorithm, and determining a global inspection path of the unmanned aerial vehicle; performing routing inspection according to the determined global routing inspection path to obtain a global routing inspection result; determining a to-be-inspected position of the climbing robot in the minimum convex polygon according to the global inspection result; and determining a vertical inspection path of the climbing robot by adopting an improved branch and bound algorithm according to the to-be-inspected position. According to the invention, more efficient, safer and more intelligent inspection operation can be realized, and the flexibility and efficiency of task execution are improved while a wider area is covered.

Description

Technical Field

[0001] The present invention relates to the technical field of robots, and particularly to a collaborative three-dimensional inspection planning method for heterogeneous robots under vertical structure facilities. Background Art

[0002] Facility inspection plays an important role in maintaining the functions of facilities. In the early applications of robot inspection, the robot-based method provides higher efficiency and flexibility than manual solutions. The robot system can get closer to the target and obtain information at a given position with higher precision, but has a limited field of view. In addition, robots are prone to encounter obstacles during the detection process, resulting in the inspection falling into a local optimum. Subsequently, researchers explored the use of unmanned aerial vehicles (UAVs) for facility-related inspections to solve this problem. UAVs are more flexible and have a better field of view than robots, and can quickly cover large areas of facilities. However, they have a smaller payload and a shorter operation time, and are vulnerable to flight time and external environmental interference, and cannot perform contact inspections. In recent years, there has been an increasing amount of research on heterogeneous robot systems. These two types of devices can complement each other to improve the overall performance of detection. However, for heterogeneous systems, the huge differences in sensor settings, mobility, viewing angles, and processing capabilities pose challenges to coordination strategies.

[0003] To address this challenge and solve the problem of insufficient coordination processing ability during the detection process of heterogeneous robot systems, the present invention proposes a collaborative three-dimensional inspection path planning method for land-air heterogeneous robots under vertical structure facilities. Summary of the Invention

[0004] To solve the problems existing in the prior art, the present invention provides a collaborative three-dimensional inspection planning method for heterogeneous robots under vertical structure facilities. The method includes: obtaining point cloud data of a target facility with a vertical structure; generating a minimum convex polyhedron describing the target facility with a vertical structure by using a three-dimensional convex hull enhancement algorithm based on the point cloud data; determining a cruise start point and a cruise end point of a UAV based on the minimum convex polyhedron; performing path planning between the cruise start point and the cruise end point of the UAV by using an improved A-star algorithm to determine a global inspection path of the UAV; performing inspection according to the determined global inspection path to obtain a global inspection result; determining positions to be inspected by a climbing robot in the minimum convex polyhedron according to the global inspection result; and determining a vertical inspection path of the climbing robot by using an improved branch and bound algorithm according to the positions to be inspected. The present invention can achieve more efficient, safer, and more intelligent inspection operations, and improve the flexibility and efficiency of task execution while covering a wider area.

[0005] The present invention adopts the following technical solutions. A collaborative three-dimensional inspection planning method for heterogeneous robots under vertical structure facilities includes: Obtain the point cloud data of the target facility with a vertical structure; generate the minimum convex polyhedron describing the vertical structure of the target facility by using a three-dimensional convex hull enhancement algorithm according to the point cloud data; Determine the cruise starting point and the cruise ending point of the unmanned aerial vehicle (UAV) based on the minimum convex polyhedron; Perform path planning between the cruise starting point and the cruise ending point of the UAV by using an improved A* algorithm to determine the global inspection path of the UAV; The UAV performs inspections according to the determined global inspection path to obtain a global inspection result; Determine the positions to be inspected by the climbing robot in the minimum convex polyhedron according to the global inspection result; According to the positions to be inspected, use an improved branch and bound algorithm to determine the vertical inspection path of the climbing robot.

[0006] Further, generating the minimum convex polyhedron of the target facility describing the vertical structure by using a three-dimensional convex hull enhancement algorithm according to the point cloud data is specifically as follows: Select three non-collinear point clouds in the point cloud data of the target facility with a vertical structure as three vertices to establish an initial convex hull; Select any point cloud in the point cloud data of the target facility with a vertical structure, and determine whether the point cloud is located inside the initial convex hull. If it is located inside, skip it; If it is not located inside, determine the positional relationship between the point cloud and each edge of the initial convex hull, and add the point cloud to the initial convex hull according to the positional relationship; Traverse all the point clouds in the point cloud data of the target facility with a vertical structure in sequence to obtain a convex hull point cloud set describing the target facility with a vertical structure; Obtain the minimum convex polyhedron of the target facility describing the vertical structure according to the convex hull point cloud set.

[0007] Further, performing path planning between the cruise starting point and the cruise ending point of the UAV by using an improved A* algorithm includes: Set a heuristic function, and the heuristic function is used to estimate the cost between the cruise starting point and the cruise ending point of the UAV; Construct a cost function according to the turning cost, rising cost and collision avoidance cost between the cruise starting point and the cruise ending point of the UAV; Start from the cruise starting point to perform the search for the next viewpoint, calculate the priority of the next viewpoint according to the heuristic function and the cost function, and select the next viewpoint with the highest priority as the optimal viewpoint; Take the optimal viewpoint as the starting point, perform the search for the next viewpoint again, and select the next viewpoint with the highest priority as the optimal viewpoint; Iterate in sequence until reaching the cruise ending point to obtain all the optimal viewpoints; Obtain the global inspection path of the UAV according to all the optimal viewpoints.

[0008] Further, when obtaining the global inspection path of the UAV based on all the optimal viewpoints, it also includes: Set the position coordinates of the previous optimal viewpoint as , and the position coordinates of the next optimal viewpoint as ; Set the UAV speed as , the optimal track angle as and the optimal heading angle as , and conduct track planning between two adjacent optimal viewpoints, expressed as: ; ; Among them, represents the objective function of the optimal track planning, , , are the velocity components of the UAV in the x, y, and z coordinate axes respectively. The integration interval is from t = 0 to t = k, indicating that the time experienced by the UAV flying from the previous optimal viewpoint to the adjacent next optimal viewpoint is k; , , are the offsets in the x, y, and z coordinate axes respectively. Here, it is the velocity attached to the upper wind of the UAV, () is the four-quadrant arctangent function; , , represent the position coordinates of the previous optimal viewpoint on the x, y, and z axes at time t = 0 , , represent the position coordinates of the next optimal viewpoint on the x, y, and z axes at time t = k, , then the optimal heading angle , and construct the system Hamiltonian function through the real-time wind speed measurement of the UAV: ; Among them, , , are the Lagrange multipliers respectively. By taking the partial derivatives of and in the system Hamiltonian function H respectively and setting them equal to 0, the optimal track angle and the optimal heading angle are obtained; Obtain the optimal track path between two adjacent optimal viewpoints according to the optimal track angle and the optimal heading angle; Obtain the optimal flight path for adjacent optimal viewpoints in the global inspection path in sequence to obtain the optimized global inspection path.

[0009] Furthermore, determine the positions to be inspected by the climbing robot in the smallest convex polyhedron according to the global inspection result, specifically: Obtain the set of inspection positions of the UAV according to the global inspection result; Obtain the climbing angle and descending angle of the climbing robot; Search for the positions to be inspected in the set of inspection positions according to the climbing angle and the descending angle to obtain the positions to be inspected by the climbing robot, and construct a set of positions to be inspected.

[0010] Furthermore, use an improved branch and bound algorithm to determine the vertical inspection path of the climbing robot, including: Generate the initial inspection path of the climbing robot using a greedy algorithm according to the set of positions to be inspected; Take the initial feasible solution of the greedy algorithm as the initial upper bound; Start from the first position to be inspected in the set of positions to be inspected, and expand task points between the next positions to be inspected; Calculate the upper bound of the search tree of task points. If the upper bound of the search tree is greater than the initial upper bound, re-expand the task points; if the upper bound of the search tree is less than the initial upper bound, take this task point as the next position to be inspected; Start from the next position to be inspected and expand task points again, and iterate in sequence to obtain the vertical inspection path.

[0011] The beneficial effects of the present invention are as follows: The inspection plan provided by the present invention includes two parts: the visual inspection plan of the UAV and the contact inspection plan of the climbing robot. Among them, the visual inspection plan of the UAV constructs a three-dimensional point cloud description of the vertical structure facility through a three-dimensional convex hull incremental algorithm, obtains the inspection viewpoints of the vertical structure facility, and completes the inspection of the UAV viewpoints based on the viewpoint description using an improved A* algorithm, thus meeting the requirements of the minimum inspection path planning; the contact inspection plan of the climbing robot is based on the inspection result of the UAV, and thus constructs a rolling branch and bound algorithm to complete the inspection. Through this algorithm, real-time inspection plan adjustment is carried out to meet the increasing inspection tasks based on the UAV, and while meeting the inspection quality, it can also meet the requirement of minimizing the inspection path. Description of the Drawings

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0013] Figure 1 Schematic diagram of the process of the collaborative three-dimensional inspection planning method for heterogeneous robots under a vertical structure facility in an embodiment of the present invention; Figure 2 Schematic diagram of the principle of viewpoint acquisition in an embodiment of the present invention; Figure 3 Schematic diagram of the process of the inspection path planning method for an unmanned aerial vehicle in an embodiment of the present invention; Figure 4 Schematic diagram of the process of the inspection path planning method for a climbing robot in an embodiment of the present invention. Detailed implementation manners

[0014] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0015] A schematic diagram of the process of the collaborative three-dimensional inspection planning method for heterogeneous robots under a vertical structure facility in an embodiment of the present invention is as Figure 1 shown and includes: Obtain the point cloud data of the target facility with a vertical structure; generate the minimum convex polyhedron describing the target facility with a vertical structure by using a three-dimensional convex hull enhancement algorithm; In the embodiment of the present invention, in order to realize the inspection of the target facility with a vertical structure, first generate a point cloud on the target facility with a vertical structure: given the projection information and height information of the target facility with a vertical structure, evenly divide the circumferential angle of the target facility with a vertical structure into 2n equal parts to obtain the azimuth angle , n = , take the centroid point as the starting point, and use the azimuth angle as to make a ray, generate a point at the boundary of the target facility with a vertical structure on each divided ray. If a given height is given, 2n points of cloud are generated at the given height. If the height of the target facility with a vertical structure is divided into m equal parts of the given height, the number of generated points of cloud is m×2n.

[0016] In an embodiment of the present invention, a three-dimensional convex hull enhancement algorithm is further used to implement the minimum convex polyhedron description of the target facility with a vertical structure. The three-dimensional convex hull enhancement algorithm can be expressed as: ; Among them, is the minimum convex polyhedron with the smallest volume to be solved, and respectively represent the volume of the selected convex polyhedron and the volume of the target facility with a vertical structure. The specific implementation steps are as follows: Select three non-collinear points p1, p2, and p3 in the point cloud as the three vertices of the initial convex hull; traverse the point cloud: for any point p in the point cloud, if the point p is inside the initial convex hull, skip it; otherwise, traverse all the edges of the initial convex hull, and determine whether the point p is outside some faces. If so, delete the face; for all the edges of the initial convex hull, check whether the point p is on one side of the edge. If so, delete the face and add two new faces with the point p and the endpoints on the edge as vertices; add the point p to the initial convex hull: when all the points in the point cloud are traversed, the three-dimensional convex hull can be obtained, and the point cloud set of the describable three-dimensional convex hull can be obtained; based on the point cloud set of the three-dimensional convex hull, the unmanned aerial vehicle in the air constructs corresponding viewpoints to create a coverage path, and the climbing robot constructs a stochastic traveling salesman problem based on the real-time changing inspection requirements to complete the inspection task.

[0017] Determine the starting point and ending point of the UAV's cruise based on the minimum convex polyhedron; In an embodiment of the present invention, based on the minimum convex polygon description of the target facility with a vertical structure, the inspection viewpoints of the target facility with a vertical structure are obtained. Specifically, in the horizontally divided sub-regions, based on the task division result, a safety distance between the UAV and the facility wall is given, and through the regional grid , the corresponding viewpoints are obtained , which is expressed as: ; Among them, represents the grid, is the normal vector of the grid , i represents the i-th grid cell, , which represents an offset used to determine the relative in the direction, d is the required sampling distance, which represents the safety distance between the UAV and the wall of the target facility with a vertical structure, is the detection width of the vision sensor installed on the UAV, is the horizontal field of view of the vision sensor installed on the UAV. A method for obtaining viewpoints in an embodiment of the present invention is as Figure 2As shown in the figure, further based on the viewpoint acquisition method, the starting point and ending point of the drone's cruise on the target facility with a vertical structure can be further determined according to the actual task requirements.

[0018] An improved A* algorithm is used to plan a path between the starting point and ending point of the drone's cruise, and the global inspection path of the drone is determined. In the embodiment of the present invention, the drone inspection uses an improved A* algorithm (i.e., the A* algorithm). This algorithm plans a path by introducing optimization measures such as collision prediction, adaptive step size adjustment, and collision threat cost function, and safely and effectively crosses the vertical structure facility environment to complete the inspection coverage. The schematic flow diagram is as Figure 3 shown. By setting a heuristic function, the heuristic function is used to estimate the cost between the starting point and ending point of the drone's cruise; a cost function is constructed based on the turning cost, rising cost, and anti-collision cost of the drone from the starting point to the ending point; starting from the starting point of the cruise, the next viewpoint is searched, and the priority of the next viewpoint is calculated according to the heuristic function and the cost function, and the next viewpoint with the highest priority is selected as the optimal viewpoint; the optimal viewpoint is used as the starting point, the next viewpoint is searched again, and the next viewpoint with the highest priority is selected as the optimal viewpoint; iterate in turn until the ending point of the cruise is reached to obtain all the optimal viewpoints; the global inspection path of the drone is obtained according to all the optimal viewpoints.

[0019] Specifically as follows: (1) Set the heuristic function: The heuristic function h(n) is used to estimate the estimated cost from the starting point of the cruise to the next viewpoint for each viewpoint. In a vertical structure, the Euclidean distance can be used as the heuristic function. The embodiment of the present invention gives a cost calculation method from the current viewpoint to the next optimal viewpoint, that is: ; Among them, is the current viewpoint, is the next optimal viewpoint, ( ) is the coordinate of the current viewpoint on the x, y, and z axes, ( ) is the coordinate of the next optimal viewpoint on the x, y, and z axes.

[0020] (2) Set the cost function: The cost function g(n) represents the actual cost from the starting point of the cruise to the next viewpoint for each viewpoint. In a vertical structure, the embodiment of the present invention constructs a cost function by considering the turning cost, rising cost, and anti-collision cost of the drone. Taking the actual cost from the current viewpoint to the next viewpoint as an example, the cost function can be expressed as: ; Among them, d represents the Euclidean distance from the current viewpoint to the next optimal viewpoint, , , are the cost coefficients respectively, ; turning cost is calculated based on the angle between the current path and the previous path, that is: ; wherein, and represent the direction vectors of the current flight segment and the next flight segment respectively. Among them, the current flight segment refers to any next view point with the current view point as the previous optimal view point. That is, the current view point may also be the next optimal view point of the previous optimal view point. Then, the path from the previous optimal view point to the current view point is the current flight segment, and the path between the current flight segment and any next view point is the next flight segment; the ascending cost is calculated based on the height difference, that is: ; safety cost is used to avoid approaching obstacles and can be calculated by the reciprocal of the distance between the current view point and the obstacle, that is ; wherein, represents the position coordinates of the current view point, represents the position coordinates of the obstacle. The closer the distance between the current view point and the obstacle is, the higher the safety cost is.

[0021] Starting from the cruise starting point, the improved A-star algorithm is used for path search. In each iteration, the comprehensive priority of each view point is calculated according to the heuristic function and the cost function, f(n)=g(n)+h(n), and the view point with the highest priority is selected for expansion. After determining the optimal view point for expansion, route planning needs to be carried out between two points. For three-dimensional inspection, the UAV inspection faces the problem of wind shear caused by the interaction between the wind field and the vertical structure, which in turn causes turbulence. Therefore, the embodiment of the present invention further gives the UAV dynamics description including the wind field description, and then corrects the inspection trajectory online. For the trajectory planning between two given view points, the position coordinates of the previous optimal view point are , and the position coordinates of the next optimal view point are . Define the UAV speed as , the optimal track angle as and the optimal heading angle as , then the optimal track planning is expressed as: ; ; wherein, represents the objective function of the optimal track planning, , , are the velocity components of the drone in the x, y, and z axis directions respectively. The integration interval is from t = 0 to t = k, indicating that the time taken for the drone to fly from the previous optimal viewpoint to the adjacent next optimal viewpoint is k; and and are the offsets in the x, y, and z axis directions respectively. Here, it is the velocity of the wind attached to the drone () is the four - quadrant arctangent function; and and represent the position coordinates on the x, y, and z axes corresponding to the previous optimal viewpoint at time t = 0 and and represent the position coordinates on the x, y, and z axes corresponding to the next optimal viewpoint at time t = k. , then the optimal heading angle , construct the system Hamiltonian function through the real - time wind speed measurement of the drone: ; Among them, and and are Lagrange multipliers respectively. By taking the partial derivatives of and in the system Hamiltonian function H respectively and setting them equal to 0, the optimal track angle and the optimal heading angle can be obtained, and then the path correction can be realized, enabling the drone to reach the next optimal viewpoint position from the initial viewpoint position in the shortest time, achieving the patrol path track planning, that is, obtaining the global patrol path of the embodiment of the present invention.

[0022] The drone conducts patrol according to the determined global patrol path to obtain the global patrol result; The drone patrol is vulnerable to external environmental influences such as wind field interference, resulting in missed detection in some areas. In addition, for specific areas with insufficient scene information in vertical facilities, such as valves, pipeline joints, and suspected rupture areas in storage tanks, climbing robots need to conduct refined patrol one by one to achieve enhanced information perception.

[0023] Therefore, in the embodiment of the present invention, the drone is used for online patrol to obtain the local surface boundary, and the climbing robot extracts the surface boundary from the shared map information. Secondly, as the drone's online patrol progresses, through information sharing, the robot will gradually add the missed detection areas, forming an uncertain patrol planning problem, thereby obtaining the global patrol result; at this time, the patrol task set consists of a group of known patrol positions and a group of uncertain patrol positions Composed of, among which there are m determined positions, represented by the set S, which are respectively , up to . Another type is the undetermined inspection positions, represented by the set , which contains n - m positions, from to .

[0024] Determine the positions to be inspected by the climbing robot in the smallest convex polyhedron according to the global inspection result; A heterogeneous robot system usually integrates multiple different types of robots and completes complex tasks that are difficult for a single type of robot to complete through collaborative work; in the embodiment of the present invention, the heterogeneous robot is a structural set of an unmanned aerial vehicle and a climbing robot. Through the collaborative inspection of the unmanned aerial vehicle and the climbing robot, precise inspection under a vertical structure is realized. Therefore, after determining the global inspection path of the unmanned aerial vehicle, it is also necessary to determine the inspection path of the climbing robot.

[0025] The embodiment of the present invention sets the initial position of the climbing robot as , and the distance between the inspection position and the inspection position is . The climbing angle and the descending angle are restricted, that is, , that is, are the minimum and maximum offset angles for climbing. The problem is constructed into a variant of the traveling salesman problem with a fixed starting point. Considering trying to visit each inspection position and each position is only inspected once, a coverage inspection path of the climbing robot based on the global inspection path can be established as: ; s.t. ; Among them, represents the total length of the coverage inspection path, represents the distance from the inspection position to the next inspection position , is the initial position of the climbing robot and also the end position, is a permutation, representing the access order of the inspection positions, represents the distance from the inspection position to the end position of the climbing robot, represents the distance from the initial position of the climbing robot to the next unvisited inspection position The distance, which represents the distance from the initial position of the climbing robot Start, obtain the distance to the next unvisited inspection position to be visited of, and sequentially obtain the next unvisited inspection position to be visited , until the last unvisited inspection position to be visited , and finally return to the end position of the climbing robot ; n is the total number of inspection positions; minimizing the maximum path length of the inspection path planning is taken as the main goal of covering the inspection. At this time, the covered inspection path contains many positions not inspected in the global inspection path, which is the inspection path to be inspected by the climbing robot.

[0026] According to the inspection positions to be inspected, an improved branch and bound algorithm is used to determine the vertical inspection path of the climbing robot.

[0027] The embodiment of the present invention uses an improved branch and bound algorithm to optimize the climbing transition gait to minimize the inspection path, and its schematic flow diagram is as Figure 4 shown, and is specifically implemented according to the following steps: (1) Initialization: Given the set of known inspection positions }, the UAV misses inspection positions during the initial process , the starting position of the climbing robot , and the minimum climbing offset angle and the maximum offset angle are given.

[0028] (2) Taking the set of known inspection positions as the inspection target, using the greedy algorithm to generate an initial path, and gradually selecting the nearest inspection position on the basis of the initial path until all known inspection positions are inspected. At this time, an initial feasible solution is generated as the current upper bound , the lower bound is obtained by using the linear programming relaxation algorithm, that is: ; ; Among them, here represents the distance from inspection position i to inspection position j, is a binary variable. When directly moving from inspection position i to inspection position j, = 1; otherwise = 0, and the goal is to minimize the total distance of all inspection paths; the simplex method is used to solve the relaxation problem, is the total number of inspection positions, and the constraint means that for each inspection position j, there is and only one path entering this inspection position, and the constraint It means that for each inspection position \(i\), there is exactly one path starting from this inspection position. That is, the first and second constraints mean that each inspection position is entered exactly once and each inspection position is left exactly once. The third constraint represents the elimination of sub-loop constraints, which means that for the values of \(i\) and \(j\) within the range of , and when \(i\neq j\), this inequality is satisfied. When , the inequality becomes ; when , the inequality becomes , that is ; in the Traveling Salesman Problem (TSP, the inspection path planning problem here can be regarded as a variant of TSP), there may be some small loops that do not include all inspection positions. This constraint condition can avoid this situation and ensure that a complete path traversing all inspection positions is obtained.

[0029] (3) The inspection task starts from the initial position of the climbing robot and gradually expands the task points. Starting from the last inspection position of the current path, considering the angular relationship, an unvisited inspection position is selected as the next unvisited inspection position to be visited , calculate the height difference between the two inspection positions , and at the same time calculate the horizontal distance between the two inspection positions: ; In the formula, represents the horizontal distance between the two inspection positions, represents the unvisited inspection position selected as the next one to be visited, represents the horizontal distance between the last inspection position \(s\) of the current path and the next unvisited inspection position , , represent the \(x\)-axis coordinate and \(y\)-axis coordinate of the last inspection position of the current path in the plane rectangular coordinate system, , represent the \(x\)-axis coordinate and \(y\)-axis coordinate of the next unvisited inspection position in the plane rectangular coordinate system.

[0030] Calculate the offset angle , check whether the angle satisfies the constraint conditions. If the path does not satisfy the angle constraint, it means that this task point is currently unreachable; if it satisfies the constraint conditions, based on the distance between the two inspection positions in the current path, taking as the next task point to calculate the upper bound of the search tree. If , pruning operation is performed to ignore the newly formed local feasible solutions; if , the upper bound is updated as ; the lower bound is calculated using the relaxation algorithm , if , pruning operation is performed to remove all local feasible solutions; iterative search is performed on all task points in turn. When the search tree traversal is completed, the path corresponding to the current upper bound is returned as the optimal path. At the same time, a smooth path planning is completed using a cubic Bezier curve. In the embodiment of the present invention, according to the surface shape and the robot motion constraints, the intermediate control points are determined, and the Bezier curve is described as: ; wherein, , represent the start point and the end point of the path point, and the intermediate control point should be in the direction of the tangent line at the start point, and the intermediate control point should be in the opposite direction of the tangent line at the end point.

[0031] ; ; wherein, k is a scaling factor that controls the degree of curve bending, is the unit direction vector of the tangent line at the start point, is the unit direction vector of the tangent line at the end point.

[0032] (4) During the inspection process, the UAV may discover new undetected areas and generate new uncertain tasks. For any newly added task point, it is added to the task set. This added task only affects the uninspected path and has no impact on the inspected path. Therefore, the greedy algorithm is used to quickly generate a new feasible solution and update the upper bound and the lower bound is recalculated using the linear programming relaxation algorithm . The inspection process continues. The climbing UAV starts from the current task point, expands the task points, still performs pruning operations considering the angle constraints, and performs further pruning operations according to the upper and lower bound calculations, and finally returns the updated path.

[0033] In the embodiment of the present invention, the improved branch and bound method can effectively solve the path planning problem of the UAV and the climbing robot during the inspection process through dynamic upper and lower bound updates, incremental search, and angle constraint integration. This method has strong dynamic adjustment ability and calculation efficiency while ensuring accuracy, and is suitable for actual application scenarios.

[0034] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A collaborative three-dimensional inspection planning method for heterogeneous robots under vertical structure facilities, characterized in that, Including: Obtaining point cloud data of a target facility with a vertical structure; Generating a minimum convex polyhedron describing the target facility with a vertical structure by using a three-dimensional convex hull enhancement algorithm based on the point cloud data; Determining the cruise starting point and the cruise ending point of the unmanned aerial vehicle (UAV) based on the minimum convex polyhedron; Performing path planning between the cruise starting point and the cruise ending point of the UAV by using an improved A* algorithm to determine the global inspection path of the UAV; The UAV performs inspections according to the determined global inspection path to obtain a global inspection result; Determining the positions to be inspected by the climbing robot in the minimum convex polyhedron according to the global inspection result; According to the positions to be inspected, using an improved branch and bound algorithm to determine the vertical inspection path of the climbing robot.

2. The collaborative three-dimensional inspection planning method for heterogeneous robots under vertical structure facilities according to claim 1, wherein: Generating a minimum convex polyhedron describing the target facility with a vertical structure by using a three-dimensional convex hull enhancement algorithm based on the point cloud data, specifically: Selecting three non-collinear point clouds in the point cloud data of the target facility with a vertical structure as three vertices to establish an initial convex hull; Selecting any point cloud in the point cloud data of the target facility with a vertical structure, and judging whether the point cloud is located inside the initial convex hull. If it is located inside, skip it; If it is not located inside, judging the positional relationship between the point cloud and each edge of the initial convex hull, and adding the point cloud to the initial convex hull according to the positional relationship; Traversing all the point clouds in the point cloud data of the target facility with a vertical structure in sequence to obtain a convex hull point cloud set describing the target facility with a vertical structure; Obtaining a minimum convex polyhedron describing the target facility with a vertical structure according to the convex hull point cloud set.

3. A method for collaborative three-dimensional inspection planning of heterogeneous robots under vertical structure facilities according to claim 1, characterized in that: Performing path planning between the cruise starting point and the cruise ending point of the UAV by using an improved A* algorithm, including: Setting a heuristic function, where the heuristic function is used to estimate the cost between the cruise starting point and the cruise ending point of the UAV; Constructing a cost function according to the turning cost, rising cost and collision avoidance cost between the cruise starting point and the cruise ending point of the UAV; Starting from the cruise starting point to perform the search for the next viewing point, calculating the priority of the next viewing point according to the heuristic function and the cost function, and selecting the next viewing point with the highest priority as the optimal viewing point; Taking the optimal viewing point as the starting point, performing the search for the next viewing point again, and selecting the next viewing point with the highest priority as the optimal viewing point; Iterating in sequence until reaching the cruise ending point to obtain all the optimal viewing points; Obtaining the global inspection path of the UAV according to all the optimal viewing points.

4. A method for collaborative three-dimensional inspection planning of heterogeneous robots under vertical structure facilities according to claim 3, characterized in that: When obtaining the global inspection path of the UAV according to all the optimal viewing points, it further includes: Set the position coordinates of the previous optimal viewpoint as , and the position coordinates of the next optimal viewpoint as ; Set the UAV speed as , the optimal track angle as and the optimal heading angle as , and conduct trajectory planning between two adjacent optimal viewpoints, expressed as: ； ； Among them, represents the objective function of the optimal trajectory planning, , , are the velocity components of the UAV in the x, y, and z axis directions respectively. The integration interval is from t = 0 to t = k, indicating that the time experienced by the UAV flying from the previous optimal viewpoint to the adjacent next optimal viewpoint is k; , , are the offsets of the wind speed attached to the UAV in the x, y, and z axis directions respectively, () is the four quadrant arctangent function; , , represent the position coordinates on the x, y, and z axes corresponding to the previous optimal viewpoint at time t = 0, , , represent the position coordinates on the x, y, and z axes corresponding to the next optimal viewpoint at time t = k. If , then the optimal heading angle ; Construct the system Hamiltonian function through real-time wind speed measurement of the UAV: ； Among them, , , are Lagrange multipliers respectively. By taking partial derivatives of and in the system Hamiltonian function H respectively and setting them equal to 0, the optimal track angle and the optimal heading angle are obtained; Obtaining the optimal track path between two adjacent optimal viewing points according to the optimal track angle and the optimal heading angle; Successively obtaining the optimal track paths for adjacent optimal viewing points in the global inspection path to obtain an optimized global inspection path.

5. A method for collaborative three-dimensional inspection planning of heterogeneous robots under vertical structure facilities according to claim 1, characterized in that: Determining the positions to be inspected by the climbing robot in the minimum convex polyhedron according to the global inspection result, specifically: Obtaining the inspection position set of the UAV according to the global inspection result; Obtaining the climbing angle and the descending angle of the climbing robot; Searching for the positions to be inspected in the inspection position set according to the climbing angle and the descending angle to obtain the positions to be inspected by the climbing robot, and constructing a set of positions to be inspected.

6. A method for collaborative three-dimensional inspection planning of heterogeneous robots under vertical structure facilities according to claim 5, characterized in that: An improved branch and bound algorithm is used to determine the vertical inspection path of the climbing robot, including: Generating an initial inspection path of the climbing robot by using a greedy algorithm according to the set of positions to be inspected; Taking the initial feasible solution of the greedy algorithm as the initial upper bound; Starting from the first position to be inspected in the set of positions to be inspected, expanding task points between the next positions to be inspected; Calculating the upper bound of the search tree of the task points. If the upper bound of the search tree is greater than the initial upper bound, re-expand the task points; if the upper bound of the search tree is less than the initial upper bound, take this task point as the next position to be inspected; Starting from the next position to be inspected, expanding task points again, and iterating in turn to obtain the vertical inspection path.

Images