Path planning method and device for unmanned aerial vehicle inspection

Abstract

The invention discloses an unmanned aerial vehicle routing inspection path planning method and device, and the method comprises the steps: obtaining a three-dimensional routing inspection map of a target routing inspection region, the three-dimensional routing inspection map comprising routing inspection equipment and obstacles; constructing a target function based on the inspection coverage rate, the inspection energy consumption and the inspection efficiency of the unmanned aerial vehicle in the target inspection area; constructing constraint conditions based on the inspection speed and the inspection time of the unmanned aerial vehicle and the shooting limitation of the unmanned aerial vehicle; based on the objective function and the constraint condition, adopting a local optimization strategy to obtain an initial planning path; and optimizing the initial planning path by adopting a global optimization strategy to complete the path planning of the unmanned aerial vehicle inspection. Through the method, automatic, efficient and accurate route planning of the inspection unmanned aerial vehicle in the target inspection area is realized, blind areas of inspection equipment are reduced to the maximum extent, the integrity and reliability of inspection are enhanced, the inspection efficiency is improved, and the operation risk is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of path planning, and in particular relates to a path planning method and device for unmanned aerial vehicle inspection. Background Art

[0002] With the rapid development of inspection drone technology, its application in the power industry has become increasingly in-depth. Drones equipped with visual sensors have become an important tool to improve inspection efficiency and reduce manual risks. For example, infrastructure inspection and power line inspection based on industrial drones. There are many facilities in substations and many large-scale basic equipment. How to cover as much infrastructure as possible during drone inspections to reduce inspection blind spots and increase inspection coverage to ensure that no inspections are missed is a problem that needs to be solved at present.

[0003] At present, most of the research related to inspection drone trajectory planning is aimed at the coverage path planning (CPP) problem. However, the CPP solution can only cover part of the target structure, and the coverage rate is low. In order to improve the coverage rate of inspection drone trajectories, researchers proposed a method of reconnaissance and then coverage, which first conducts reconnaissance and establishes a rough initial map based on the CPP method, and then forms a continuous drone trajectory based on the initial map and adopts the Traveling Salesman Problem (TSP)-3D tensor parallelism (TP) method. For large-scale drone trajectory planning problems, most researchers regard metaheuristic algorithms as solution algorithms, discretizing the entire target structure in advance, generating a large number of triangular or rectangular grids, and calculating the viewpoints corresponding to each grid. This method is very computationally intensive.

[0004] Patent CN119104052A discloses a method for multi-target UAV path planning, including: collecting geographic information data of the inspection area, establishing a planning path database based on the collected data, and preprocessing and optimizing the data; solving the UAV path planning model based on the particle swarm algorithm according to the maneuverability characteristics, safety constraints and inspection target priorities of the UAV flight; real-time detection of unconventional obstacles based on the UAV visual system to achieve local safety adjustment of the path; updating the path and sending it to the remote client synchronously for path evaluation and manual intervention. The solution comprehensively considers the flight time, path smoothness and path effectiveness, can adapt to complex environments and inspection needs, and comprehensively applies the particle swarm algorithm and binocular ranging technology to improve the intelligence and real-time performance of path planning, reduce flight risks, and ensure the safety of UAV inspection needs.

[0005] Therefore, there are still technical bottlenecks in the formulation of UAV flight paths that fully cover inspection targets. How to make the UAV's path planning adapt to the complex environment of the substation to ensure the efficient and accurate completion of the inspection mission is a problem that needs to be solved. Summary of the invention

[0006] In view of the defects in the above-mentioned prior art, the present invention provides a path planning method for unmanned aerial vehicle inspection, comprising:

[0007] Obtaining a three-dimensional inspection map of the target inspection area, wherein the three-dimensional inspection map includes inspection equipment and obstacles;

[0008] Construct an objective function based on the inspection coverage, inspection energy consumption, and inspection efficiency of the UAV in the target inspection area;

[0009] Construct constraints based on the inspection speed and time of the drone and the photography restrictions of the drone;

[0010] Based on the objective function and constraints, a local optimization strategy is adopted to obtain the initial planning path;

[0011] A global optimization strategy is adopted to optimize the initial planned path and complete the path planning of UAV inspection.

[0012] In a possible implementation, based on the inspection coverage rate, inspection energy consumption, and inspection efficiency of the UAV in the target inspection area, an objective function is constructed, specifically including:

[0013] Based on the inspection power and inspection time of the UAV in the target inspection area and the impact of the environment on the UAV, the inspection energy consumption is determined;

[0014] Based on the division of the target inspection area and the inspection path of the drone in the target inspection area, the inspection coverage rate is given;

[0015] Determine the inspection efficiency of drones by combining inspection coverage and inspection time;

[0016] The impact of inspection energy consumption, inspection coverage and inspection efficiency on UAV inspection is analyzed, and the objective function is constructed.

[0017] In a possible implementation manner, the inspection energy consumption is specifically expressed as:

[0018]

[0019] Among them, f energy is the inspection energy consumption, T i is the inspection time of the UAV on the trajectory segment from trajectory point i-1 to trajectory point i, F iis the thrust required for the UAV to travel from trajectory point i-1 to trajectory point i, δ is the number of rotors of the UAV, ρ is the air density, It is the area of ​​a single rotor of the UAV.

[0020] In a possible implementation manner, the inspection efficiency is specifically expressed as:

[0021]

[0022] Among them, f time is the total flight time of the UAV in the target inspection area, and the detection efficiency f eff is the inspection efficiency of the UAV in the target inspection area, T i is the inspection time of the UAV on the trajectory segment from trajectory point i-1 to trajectory point i, Γ is the set of trajectory points in the three-dimensional inspection map, and f cover is the inspection coverage rate.

[0023] In a possible implementation, the objective function is specifically expressed as:

[0024]

[0025] Among them, minf obj is the objective function, β1 is the influence coefficient of inspection energy consumption, β2 is the influence coefficient of inspection efficiency, β3 is the influence coefficient of inspection coverage, and f energy is the inspection energy consumption, f time is the total flight time of the UAV in the target inspection area, f cover is the inspection coverage rate.

[0026] In a possible implementation, the constraint conditions include inspection speed constraint, inspection time constraint and shooting limit constraint;

[0027] Based on the inspection speed and time of the drone and the shooting restrictions of the drone, the constraints are constructed, including:

[0028] According to the position change and attitude adjustment of the UAV, the inspection speed is given and the inspection speed constraint is constructed;

[0029] Based on the inspection time of the drone, the inspection time constraints are given;

[0030] Based on the shooting parameters of the drone's visual sensor and the angle limit of the drone's gimbal, the shooting restriction constraints are constructed.

[0031] In a possible implementation manner, the inspection speed constraint is specifically expressed as:

[0032]

[0033] Among them, X i is the change in coordinate value on the x-axis from trajectory point i-1 to trajectory point i, Y i is the change in coordinate value on the y-axis from trajectory point i-1 to trajectory point i, Z i is the change in coordinate value on the z-axis from trajectory point i-1 to trajectory point i, T i is the inspection time of the UAV on the trajectory segment from trajectory point i-1 to trajectory point i, T Ψi is the time required for the UAV to adjust its attitude at trajectory point i, is the maximum flight speed of the UAV in the horizontal direction, It is the maximum flight speed of the UAV in the vertical direction.

[0034] In a possible implementation manner, the shooting restriction constraint is specifically expressed as:

[0035]

[0036] Among them, Q i,n =(x i,n ,y i,n ,z i,n ), Q i,n is the coordinate of viewpoint n on the i-th inspection device and n∈N i , D max is the maximum shooting distance of the visual sensor on the drone, D min The minimum shooting distance of the visual sensor on the drone, α max is the maximum field of view of the visual sensor on the drone, p i is the coordinate of the i-th trajectory point in the target inspection area.

[0037] In a possible implementation, based on the objective function and the constraint conditions, a local optimization strategy is adopted to obtain an initial planning path, specifically including:

[0038] Determine the starting point from the trajectory points of the target inspection area;

[0039] Based on the objective function and constraints, combined with the distance between two trajectory points, a list of candidate points corresponding to the current trajectory point is given;

[0040] According to the candidate point list corresponding to the current trajectory point, the next trajectory point is determined in sequence to obtain an initial planning path, wherein the initial planning path includes all trajectory points in the target inspection area.

[0041] In a possible implementation manner, the initial planned path is specifically expressed as:

[0042]

[0043] Where Distance is the total distance of the initial planned path, d(c i ,c i+1 ) is the distance from trajectory point i to trajectory point i+1, c1 is the starting point of the initial planning path, c I The end point of the initial planned path.

[0044] In a possible implementation, a global optimization strategy is adopted to optimize the initial planned path to complete the path planning of the drone inspection, specifically including:

[0045] Initialize nonlinear convergence factor and strategy selection threshold;

[0046] Based on the fitness value of each initial planned path, determine the path to be optimized from the initial planned path;

[0047] Determine the optimization strategy of the path to be optimized according to the fitness value of the path to be optimized and the strategy selection threshold;

[0048] For each path to be optimized, based on the random function and the optimization strategy of the path to be optimized, the relationship between the trajectory points in the path to be optimized is adjusted to give the current planned path;

[0049] According to the current number of iterations corresponding to the current planned path, the current planned path is smoothly optimized to complete the path planning of the drone inspection.

[0050] In a possible implementation, the optimization strategy of the path to be optimized includes a differential operator strategy and a discoverer strategy;

[0051] For each path to be optimized, based on the random function and the optimization strategy of the path to be optimized, the relationship between the trajectory points in the path to be optimized is adjusted to give the current planned path, including:

[0052] Based on the random function, an optimized random number is obtained;

[0053] According to the optimization strategy and optimization random number of the path to be optimized, the initial planning path is selected, and the relationship between each trajectory point in the path to be optimized is updated based on the selected initial planning path to give the current planning path.

[0054] The present invention also discloses a path planning device for unmanned aerial vehicle inspection, comprising:

[0055] A map acquisition module is used to acquire a three-dimensional inspection map of the target inspection area, wherein the three-dimensional inspection map includes inspection equipment and obstacles;

[0056] A function building module is used to build an objective function based on the inspection coverage, inspection energy consumption and inspection efficiency of the UAV in the target inspection area;

[0057] The constraint building module is used to build constraints based on the inspection speed and inspection time of the drone and the shooting restrictions of the drone;

[0058] The path generation module is used to obtain the initial planning path based on the objective function and constraints by adopting a local optimization strategy;

[0059] The path optimization module is used to optimize the initial planned path using a global optimization strategy to complete the path planning for drone inspections.

[0060] The present invention provides a method and device for planning a path for unmanned aerial vehicle inspection, which has at least the following beneficial effects:

[0061] (1) By obtaining a three-dimensional inspection map of the target inspection area including inspection equipment and obstacles, the inspection coverage, inspection energy consumption and inspection efficiency of the UAV in the target inspection area are analyzed, and the objective function is constructed. The inspection speed, inspection time and shooting restrictions of the UAV are combined to construct constraint conditions. A two-stage analysis combining local optimization strategy and global optimization strategy is adopted to obtain the initial planning path and optimize it, and complete the path planning of the UAV inspection to ensure that the inspection UAV can automatically, efficiently and accurately plan the trajectory in the target inspection area, minimize the blind spots of the inspection equipment, enhance the integrity and reliability of the inspection, and at the same time, improve the inspection efficiency and significantly reduce the operational risk.

[0062] (2) By updating the parameters in the three-dimensional inspection map and constraint conditions, the target inspection path can be dynamically adjusted to cope with environmental changes and equipment status updates, making the inspection drone more flexible and adaptive. This improves the quality and safety of the inspection, ensures the efficient operation of the inspection equipment, and reduces the downtime and maintenance costs caused by inspection equipment failures.

[0063] (3) By considering the inspection coverage, energy consumption, and efficiency of the inspection drone, the energy consumption and hardware limitations of the drone are reduced, achieving full coverage of the inspection equipment and efficient use of the drone's energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] Figure 1 A flow chart of a method for planning a path for an unmanned aerial vehicle inspection provided by an embodiment of the present invention;

[0065] Figure 2 A flowchart of constructing an objective function in a path planning method for unmanned aerial vehicle inspection provided by an embodiment of the present invention;

[0066] Figure 3 A flowchart of constructing constraint conditions in a path planning method for unmanned aerial vehicle inspection provided by an embodiment of the present invention;

[0067] Figure 4 A flow chart of determining an initial planned path in a path planning method for unmanned aerial vehicle inspection provided in an embodiment of the present invention;

[0068] Figure 5 A flow chart for optimizing an initial planned path in a path planning method for unmanned aerial vehicle inspection provided in an embodiment of the present invention;

[0069] Figure 6 A flow chart of a method for planning a path for an unmanned aerial vehicle inspection provided by an embodiment of the present invention;

[0070] Figure 7 A structural block diagram of a path planning device for drone inspection provided in an embodiment of the present invention.

[0071] Among them, 201 is a map acquisition module; 202 is a function construction module; 203 is a constraint construction module; 204 is a path generation module; 205 is a path optimization module. DETAILED DESCRIPTION

[0072] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0073] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings, and "multiple" generally includes at least two.

[0074] It should also be noted that the term "includes", "comprising" or any other variation thereof is intended to cover non-exclusive inclusion, so that a commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprising a ..." do not exclude the existence of other identical elements in the commodity or device including the elements.

[0075] In order to solve the problems of traditional manual inspection, the use of drones equipped with visual sensors for autonomous inspection of power facilities has become a promising solution. In order to ensure the practicality of drone inspection operations, it is necessary to consider the limited durability of drone hardware and the impact of the limitations of visual sensors and gimbals on the quality of collected images. In addition, drones need to cover as much infrastructure as possible during the inspection process to reduce blind spots and increase the inspection coverage to ensure that no inspections are missed. Based on this, it is essential to be able to autonomously formulate a flight path that covers the entire target infrastructure.

[0076] At the same time, although some researchers have established multi-objective mathematical models for UAV inspection trajectory planning, the objective functions of most UAV trajectory planning mathematical models focus on reducing trajectory costs including time and distance costs, and no further research has been conducted on trajectory planning issues that consider UAV energy consumption.

[0077] In view of the limitations of related technologies, the present invention proposes a path planning method for drone inspection. The method comprehensively considers drone energy consumption, inspection efficiency and coverage of inspection equipment to construct a mathematical model for trajectory planning, effectively overcoming the limitations of drone energy consumption and drone hardware layout, and achieving full coverage of the target inspection area and efficient use of drone energy consumption. At the same time, the present invention proposes a two-stage heuristic algorithm. The first stage optimizes the set of viewpoints in the target inspection area, and the second stage generates trajectories connecting viewpoints to maximize the coverage of the inspection equipment and reduce energy and time consumption, thereby realizing autonomous trajectory planning of the drone. When the starting point and end point of the drone are given, a new set of viewpoints is found in each iteration. Unlike the previous technology that discretizes the entire inspection equipment in advance, generates a large number of triangular or rectangular grids, and calculates the viewpoints corresponding to each grid, which has a very large amount of calculation, the two-stage heuristic algorithm has a smaller amount of calculation, and at the same time has fewer detection blind spots for the inspection equipment. By integrating UAV energy consumption, visual sensor layout constraints and heuristic search methods, the algorithm can generate trajectory planning with the goal of maximizing the coverage of inspection equipment, ensuring that the UAV can conduct comprehensive inspections of inspection equipment while avoiding obstacles and no-fly zones, reducing collision and interference risks, avoiding damage to UAVs and substation equipment, and improving the safety and reliability of UAV inspections in complex electromagnetic environments.

[0078] like Figure 1 As shown, the present invention proposes a path planning method for unmanned aerial vehicle inspection, and the specific steps of the method are as follows:

[0079] S101: Acquire a three-dimensional inspection map of a target inspection area.

[0080] Specifically, the three-dimensional inspection map includes inspection equipment and obstacles. A three-dimensional inspection map is constructed based on the actual inspection range, and the drone performs the inspection task of the inspection equipment within the known and limited inspection range corresponding to the three-dimensional inspection map. A three-dimensional inspection map including the drone's flight space, inspection equipment and obstacles is established based on the map data corresponding to the inspection range. Then, the surface of the inspection equipment is discretized using a rectangular network to obtain the viewpoint set N of the i-th inspection equipment. i ={1 i ,2 i ,…,(n-1) i ,n i}, I is the number of inspection devices in the target inspection area, n i is the nth viewpoint on the i-th inspection device, the viewpoint set is the set of diagonal vertices constituting the surface rectangular grid of the inspection device, the set of trajectory points in the three-dimensional inspection map is Γ={1,2,…,I}, and each trajectory point corresponds to an inspection device.

[0081] S102: Construct an objective function based on the inspection coverage, inspection energy consumption, and inspection efficiency of the UAV in the target inspection area.

[0082] Reference Figure 2 Specifically, based on the inspection power and inspection time of the UAV in the target inspection area, combined with the impact of the environment on the UAV, the inspection energy consumption is determined. Based on the division of the target inspection area and the inspection path of the UAV in the target inspection area, the inspection coverage rate is given. Combined with the inspection coverage rate and inspection time, the inspection efficiency of the UAV is determined. The impact of inspection energy consumption, inspection coverage rate and inspection efficiency on UAV inspection is analyzed, and the objective function is constructed.

[0083] Furthermore, the inspection energy consumption is specifically expressed as:

[0084]

[0085] Among them, T i is the inspection time of the UAV on the trajectory segment from trajectory point i-1 to trajectory point i, F i is the thrust required for the UAV to move from trajectory point i-1 to trajectory point i, δ is the number of rotors of the UAV, ρ is the air density, φ blade It is the area of ​​a single rotor of the UAV.

[0086]

[0087] in, is the flight velocity vector of the UAV at trajectory point i, is the wind speed vector of the wind speed at trajectory point i, is the area projection of the drone, C d is the drag coefficient of the drone, m is the mass of the drone, g is the acceleration due to gravity, is the component of the flight velocity vector of the UAV at trajectory point i in the x direction, is the component of the flight velocity vector of the UAV at trajectory point i in the y direction, is the component of the flight velocity vector of the UAV at trajectory point i in the z direction, is the component of the wind speed vector in the x direction at trajectory point i, The component of the wind speed vector in the y direction at trajectory point i, The component of the wind speed vector in the z direction at the trajectory point i. The area projection of a drone refers to the area occupied by the projection of the drone on the plane perpendicular to the flight direction. Generally speaking, the projection area can be the swept area formed by the rotation of the drone blades.

[0088]

[0089] Among them, T i is the inspection time of the UAV from trajectory point i-1 to trajectory point i on the corresponding trajectory segment, X i is the change in coordinate value on the x-axis from trajectory point i-1 to trajectory point i, Y i is the change in coordinate value on the y-axis from trajectory point i-1 to trajectory point i, Z i is the change in coordinate value from trajectory point i-1 to trajectory point i on the z-axis, ||·|| is the Euclidean norm, is the horizontal flight speed of the UAV on the trajectory segment corresponding to the trajectory point i-1 to the trajectory point i, is the vertical flight speed of the UAV on the trajectory segment corresponding to the trajectory point i-1 to the trajectory point i, Ψ i is the angular velocity of the UAV at the i-th trajectory point, Ψ i-1 The angular velocity of the drone at the i-1th trajectory point is is the average rotation angular velocity of the drone. In a specific example, the above Take 0.8 times the maximum flight speed of the drone in the horizontal direction. It is 0.8 times the maximum flight speed of the drone in the vertical direction. is 0.8 times the maximum rotational angular velocity of the drone. In other embodiments, It can be adjusted according to actual conditions and there is no limitation on this.

[0090] Furthermore, the inspection coverage rate is specifically expressed as:

[0091]

[0092] Among them, f cover is the inspection coverage rate, K i,n is a binary variable. If the drone can completely collect the viewpoint set N at the trajectory point i i , then K i,n =1, otherwise, if the drone cannot completely collect the viewpoint set N at the trajectory point i i , then K i,n =0.

[0093] Furthermore, the inspection efficiency is specifically expressed as:

[0094]

[0095] Among them, f time is the total flight time of the UAV in the target inspection area, and the detection efficiency f eff It is the inspection efficiency of the UAV in the target inspection area.

[0096] Furthermore, the objective function is specifically expressed as:

[0097]

[0098] Among them, minf obj is the objective function, β1 is the influence coefficient of inspection energy consumption, β2 is the influence coefficient of inspection efficiency, β3 is the influence coefficient of inspection coverage, and the objective function is to minimize f under given constraints. energy , maximize f eff and maximize f cover .

[0099] S103: Construct constraint conditions based on the inspection speed and inspection time of the drone and the shooting limit of the drone.

[0100] Reference Figure 3 Specifically, the constraints include inspection speed constraint, inspection time constraint and shooting limit constraint. According to the position change and attitude adjustment of the UAV, the inspection speed is given and the inspection speed constraint is constructed. Based on the inspection time of the UAV, the inspection time constraint is given. Based on the shooting parameters of the UAV visual sensor and the angle limit of the UAV gimbal, the shooting limit constraint is constructed.

[0101] Furthermore, the inspection speed constraint is specifically expressed as:

[0102]

[0103] Among them, X i is the change in coordinate value on the x-axis from trajectory point i-1 to trajectory point i, Yi is the change in coordinate value on the y-axis from trajectory point i-1 to trajectory point i, Z i is the change in coordinate value on the z-axis from trajectory point i-1 to trajectory point i, T i is the inspection time of the UAV on the trajectory segment from trajectory point i-1 to trajectory point i, T Ψi is the time required for the UAV to adjust its attitude at trajectory point i, is the maximum flight speed of the UAV in the horizontal direction, It is the maximum flight speed of the UAV in the vertical direction.

[0104] Furthermore, the inspection time constraint is specifically expressed as:

[0105]

[0106] in, is the total flight time of the UAV in the target inspection area, T max The maximum flight time of the drone.

[0107] Furthermore, the shooting restriction constraint is specifically expressed as:

[0108]

[0109] Among them, Q i,n =(x i,n ,y i,n ,z i,n ), Q i,n is the coordinate of viewpoint n on the i-th inspection device and n∈N i , D max is the maximum shooting distance of the visual sensor on the drone, D min The minimum shooting distance of the visual sensor on the drone, α max is the maximum field of view of the visual sensor on the drone, p i is the coordinate of the i-th trajectory point in the target inspection area.

[0110] In the embodiment provided by the present invention, the constraint conditions also include an oblique angle constraint and an obstacle avoidance constraint. The oblique angle constraint is specifically expressed as:

[0111] tan(-φ max )≤Z i

[0112]

[0113] Among them, φ maxis the maximum oblique angle of the UAV on the trajectory segment corresponding to the trajectory point i-1 to the trajectory point i. The oblique angle is the track angle of the UAV, which refers to the angle between the flight trajectory of the UAV and the horizontal plane. The horizontal plane is the xoy plane, and the z-axis direction is the direction perpendicular to the horizontal plane pointing to the sky.

[0114] The obstacle avoidance constraint is specifically expressed as:

[0115]

[0116] Among them, E(i-1,i) is the edge between trajectory points i and i-1, V allobs is the set of obstacles in the target inspection area. The obstacle avoidance constraint means that for any trajectory point i, there is an edge E(i-1,i) between trajectory point i-1 and trajectory point i, and the intersection of all obstacles in the target inspection area is empty.

[0117] S104: Based on the objective function and the constraints, a local optimization strategy is adopted to obtain an initial planning path.

[0118] Reference Figure 4 Specifically, a starting point is determined from the trajectory points of the target inspection area. Based on the objective function and the constraint conditions, combined with the distance between two trajectory points, a list of candidate points corresponding to the current trajectory point is given. According to the list of candidate points corresponding to the current trajectory point, the next trajectory point is determined in turn to obtain an initial planning path, wherein the initial planning path includes all trajectory points in the target inspection area.

[0119] Furthermore, the initial planning path is specifically expressed as:

[0120]

[0121] Where Distance is the total distance of the initial planned path, d(c i ,c i+1 ) is the distance from trajectory point i to trajectory point i+1, c1 is the starting point of the initial planning path, c I The end point of the initial planned path.

[0122] In a specific implementation, the initial planning path is planned by rapidly-exploring random tree (RRT). First, the viewpoint configuration is initialized, the cost matrix of the TSP problem is calculated and the initial planning path is solved. In each iteration, the viewpoint configuration is resampled, the cost matrix is ​​recalculated and the optimal path is recalculated using the Lin-Kernighan-Helsgaun (LKH) heuristic algorithm.

[0123] It should be understood that the TSP problem can be described as given a set of cities and the distance between each pair of cities, a traveling salesman needs to start from a certain city, visit each city once, and finally return to the starting city. The goal of the traveling salesman is to find a route with the shortest total distance, that is, the travel route that he visits all cities and returns to the starting point is the shortest. In the present invention, the TSP problem can be described as that there are multiple inspection devices in the target inspection area and the distances between each inspection device are known. The drone needs to start from a certain inspection device, pass through each inspection device once, and finally return to the starting position. The drone needs to find a route with the shortest inspection distance.

[0124] Input inspection equipment set C = {c1,c2,...,c I} and the distance d(c i ,c j ), i and j are the indexes of the inspection equipment, and the distances between the trajectory points are asymmetric, that is, d(c i ,c j )≠d(c j ,c i ). The goal is to find a cycle, i.e. a path that contains each trajectory point once and returns to the starting point, such that the total distance is minimized.

[0125] The cost matrix D is an I×I matrix, where D[i][j] represents the inspection cost from inspection device i to inspection device j. In the present invention, the inspection cost is the distance between two trajectory points.

[0126] The LKH heuristic algorithm gradually optimizes the initial planning path based on dynamic k-opt operations. Determine a starting point and select the unvisited trajectory points with the shortest distance in turn until all trajectory points are visited to form a complete loop and obtain the initial planning path. The initial planning path is specifically expressed as:

[0127]

[0128] Where Distance is the total distance of the initial planned path, d(c i ,c i+1 ) is the distance from trajectory point i to trajectory point i+1, c1 is the starting point of the initial planning path, c I The end point of the initial planned path.

[0129] In other implementations, in order to reduce the amount of calculation, a candidate point list is used to limit the selection range in path improvement. The candidate list generation method is neighbor selection, which is specifically expressed as:

[0130] CandidateSet(c i )={c j|d(c i ,c j ) is the smallest k distances}

[0131] Where k represents the number of points in the candidate point list. When k = 1, it means there is only one candidate point in the candidate point list. When k = 2, it means there are two candidate points in the candidate point list. The optimization of the initial planning path is completed through dynamic k-pot operation. Starting from 2-pot, the k value is gradually increased, k edges are removed, and the path is optimized by replacing new k edges.

[0132] S105: Adopt a global optimization strategy to optimize the initial planned path and complete the path planning of the drone inspection.

[0133] Reference Figure 5 Specifically, the nonlinear convergence factor and the strategy selection threshold are initialized; based on the fitness value of each initial planned path, the path to be optimized is determined from the initial planned path; according to the fitness value and the strategy selection threshold of the path to be optimized, the optimization strategy of the path to be optimized is determined; for each path to be optimized, based on the random function and the optimization strategy of the path to be optimized, the relationship between the trajectory points in the path to be optimized is adjusted to give the current planned path; according to the current number of iterations corresponding to the current planned path, the current planned path is smoothly optimized to complete the path planning of the UAV inspection.

[0134] Furthermore, the optimization strategy of the path to be optimized includes a differential operator strategy and a discoverer strategy. Based on the random function, an optimization random number is obtained; according to the optimization strategy and the optimization random number of the path to be optimized, the initial planning path is selected, and the relationship between each trajectory point in the path to be optimized is updated based on the selected initial planning path to give the current planning path.

[0135] In a specific implementation, a nonlinear convergence factor and a strategy selection threshold are used. A safe and collision-free feasible path is generated based on the RRT algorithm and used as a set of initial planning paths. The nonlinear convergence factor is used to increase the range of the search step size to solve the problem that when a black hole reaches the optimal universe through a wormhole in a multiverse optimizer (MVO), the parameters are randomly updated in the positive and negative directions, making the global search ability of the algorithm poor.

[0136] The nonlinear convergence factor is specifically expressed as:

[0137]

[0138] Among them, CO is the nonlinear convergence factor, h is the current iteration number, and H is the maximum iteration number.

[0139] After the initialization of the parameters is completed, calculate each path B in the initial planned path o 's fitness value f(B o ), normalize the fitness values of each path B o to obtain the normalized fitness value

[0140] Sort the initial planned paths according to the total distance of each path in the initial planned path. The path with the shortest total path is used as the current optimal solution, and the optimal solution is saved as B best .

[0141] Generate a path selection random number r1∈[0,1] based on a random function. The path selection random number is used to select the path to be optimized from the initial planned path. If then do not optimize the current optimal solution. If then generate a strategy random number r2∈[0,1].

[0142] If r2 < WEP, use the differential operator strategy to optimize the path to be optimized. The specific process is as follows:

[0143] Generate an optimization random number r3∈[0,1]. The optimization formula is:

[0144]

[0145] where dv_b io is the i-th path region in the path to be optimized updated based on the differential operator strategy, is the i-th path region in the optimal solution, l max and l min respectively represent the upper and lower limits of the search space, μ i is a random number between 0 and 1, is the i-th path region of three different paths randomly selected from the initial planned path, and WEP is the strategy selection threshold. The calculation formula is as follows:

[0146]

[0147] In the embodiments of the present invention, the above path region includes at least one trajectory point in the path to be optimized.

[0148] If r2≥WEP, use the discoverer strategy to optimize the path to be optimized. Specifically, generate an optimization random number r4∈[0,1], and the update formula is:

[0149]

[0150] where b besta and bbestb is the i-th path region of the second and third best solutions in the initial planning path after sorting, and r4 and r5 are random numbers between 0 and 1, respectively.

[0151] Calculate the updated current planning path, which is specifically expressed as:

[0152]

[0153] in, It is the i-th path area in the path to be optimized after being updated based on the finder's search strategy.

[0154] Determine whether the maximum number of iterations has been reached. If so, stop the optimization process and get the current planned path. Otherwise, continue to calculate the fitness values ​​of each path for optimization. Use the Bayesian curve to smoothly optimize the current planned path and finally output a smooth three-dimensional trajectory, which is the target planned path.

[0155] Reference Figure 6 The present invention proposes a path planning method for drone inspection. The method obtains a three-dimensional inspection map of the target inspection area including inspection equipment and obstacles, analyzes the inspection coverage, inspection energy consumption and inspection efficiency of the drone in the target inspection area, constructs an objective function, combines the inspection speed, inspection time and shooting restrictions of the drone, constructs constraint conditions, and adopts a two-stage analysis combining local optimization strategy and global optimization strategy to obtain the initial planning path and optimize it to complete the path planning of drone inspection. In the process of adopting the global optimization strategy, first initialize the nonlinear convergence factor and strategy selection threshold; based on the fitness value of each initial planning path, determine the path to be optimized from the initial planning path; according to the fitness value and strategy selection threshold of the path to be optimized, determine the optimization strategy of the path to be optimized and adjust the relationship between each trajectory point in the path to be optimized to give the current planning path; according to the current number of iterations corresponding to the current planning path, smoothly optimize the current planning path to complete the path planning of drone inspection. This method can ensure that the inspection UAV can automatically, efficiently and accurately plan the trajectory for the inspection equipment in the target inspection area, and provide decision support for the inspection path of the UAV to minimize the blind spots of the inspection equipment, enhance the integrity and reliability of the inspection, and improve the inspection efficiency.

[0156] Reference Figure 7 , an embodiment of the present invention provides a path planning device for unmanned aerial vehicle inspection, comprising:

[0157] A map acquisition module 201 is used to acquire a three-dimensional inspection map of a target inspection area, wherein the three-dimensional inspection map includes inspection equipment and obstacles;

[0158] Function construction module 202, used to construct an objective function based on the inspection coverage, inspection energy consumption and inspection efficiency of the UAV in the target inspection area;

[0159] The constraint construction module 203 is used to construct constraint conditions based on the inspection speed and inspection time of the drone and the shooting limit of the drone;

[0160] A path generation module 204 is used to obtain an initial planning path based on the objective function and the constraint conditions by adopting a local optimization strategy;

[0161] The path optimization module 205 is used to optimize the initial planned path by adopting a global optimization strategy to complete the path planning of the UAV inspection.

[0162] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the described module can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0163] Although preferred embodiments of the present invention have been described, additional changes and modifications may be made to these embodiments by those skilled in the art once the basic inventive concepts are known. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A path planning method for unmanned aerial vehicle inspection, characterized in that: include: Obtaining a three-dimensional inspection map of the target inspection area, wherein the three-dimensional inspection map includes inspection equipment and obstacles; Construct an objective function based on the inspection coverage, inspection energy consumption, and inspection efficiency of the UAV in the target inspection area; Construct constraints based on the inspection speed and time of the drone and the photography restrictions of the drone; Based on the objective function and constraints, a local optimization strategy is adopted to obtain the initial planning path; A global optimization strategy is adopted to optimize the initial planned path and complete the path planning of UAV inspection.

2. The path planning method for unmanned aerial vehicle inspection according to claim 1, characterized in that: Based on the inspection coverage, inspection energy consumption and inspection efficiency of the UAV in the target inspection area, an objective function is constructed, including: Based on the inspection power and inspection time of the UAV in the target inspection area and the impact of the environment on the UAV, the inspection energy consumption is determined; Based on the division of the target inspection area and the inspection path of the drone in the target inspection area, the inspection coverage rate is given; Determine the inspection efficiency of drones by combining inspection coverage and inspection time; The impact of inspection energy consumption, inspection coverage and inspection efficiency on UAV inspection is analyzed, and the objective function is constructed.

3. The path planning method for unmanned aerial vehicle inspection as claimed in claim 2, characterized in that: The inspection energy consumption is specifically expressed as: Among them, f energy is the inspection energy consumption, T i is the inspection time of the UAV on the trajectory segment from trajectory point i-1 to trajectory point i, F i is the thrust required for the UAV to travel from trajectory point i-1 to trajectory point i, δ is the number of rotors of the UAV, ρ is the air density, It is the area of ​​a single rotor of the UAV.

4. The path planning method for unmanned aerial vehicle inspection as claimed in claim 2, characterized in that: The objective function is specifically expressed as: Among them, minf obj is the objective function, β1 is the influence coefficient of inspection energy consumption, β2 is the influence coefficient of inspection efficiency, β3 is the influence coefficient of inspection coverage, and f energy is the inspection energy consumption, f time is the total flight time of the UAV in the target inspection area, f cover is the inspection coverage rate.

5. The path planning method for unmanned aerial vehicle inspection according to claim 1, characterized in that: Constraints include inspection speed constraints, inspection time constraints, and shooting limit constraints; Based on the inspection speed and time of the drone and the shooting restrictions of the drone, the constraints are constructed, including: According to the position change and attitude adjustment of the UAV, the inspection speed is given and the inspection speed constraint is constructed; Based on the inspection time of the drone, the inspection time constraints are given; Based on the shooting parameters of the drone's visual sensor and the angle limit of the drone's gimbal, the shooting restriction constraints are constructed.

6. The path planning method for unmanned aerial vehicle inspection as claimed in claim 5, characterized in that: The inspection speed constraint is specifically expressed as: Among them, X i is the change in coordinate value on the x-axis from trajectory point i-1 to trajectory point i, Y i is the change in coordinate value on the y-axis from trajectory point i-1 to trajectory point i, Z i is the change in coordinate value on the z-axis from trajectory point i-1 to trajectory point i, T i is the inspection time of the UAV on the trajectory segment from trajectory point i-1 to trajectory point i, T Ψi is the time required for the UAV to adjust its attitude at trajectory point i, is the maximum flight speed of the UAV in the horizontal direction, It is the maximum flight speed of the UAV in the vertical direction.

7. The path planning method for unmanned aerial vehicle inspection according to claim 1, characterized in that: Based on the objective function and constraints, a local optimization strategy is adopted to obtain the initial planning path, including: Determine the starting point from the trajectory points of the target inspection area; Based on the objective function and constraints, combined with the distance between two trajectory points, a list of candidate points corresponding to the current trajectory point is given; According to the candidate point list corresponding to the current trajectory point, the next trajectory point is determined in sequence to obtain an initial planning path, wherein the initial planning path includes all trajectory points in the target inspection area.

8. The path planning method for unmanned aerial vehicle inspection as claimed in claim 1, characterized in that: The global optimization strategy is used to optimize the initial planned path and complete the path planning of the drone inspection, including: Initialize nonlinear convergence factor and strategy selection threshold; Based on the fitness value of each initial planned path, determine the path to be optimized from the initial planned path; Determine the optimization strategy of the path to be optimized according to the fitness value of the path to be optimized and the strategy selection threshold; For each path to be optimized, based on the random function and the optimization strategy of the path to be optimized, the relationship between the trajectory points in the path to be optimized is adjusted to give the current planned path; According to the current number of iterations corresponding to the current planned path, the current planned path is smoothly optimized to complete the path planning of the drone inspection.

9. The path planning method for unmanned aerial vehicle inspection as claimed in claim 8, characterized in that: The optimization strategies for the path to be optimized include differential operator strategy and discoverer strategy; For each path to be optimized, based on the random function and the optimization strategy of the path to be optimized, the relationship between the trajectory points in the path to be optimized is adjusted to give the current planned path, including: Based on the random function, an optimized random number is obtained; According to the optimization strategy and optimization random number of the path to be optimized, the initial planning path is selected, and the relationship between each trajectory point in the path to be optimized is updated based on the selected initial planning path to give the current planning path.

10. A path planning device for unmanned aerial vehicle inspection, characterized in that: A path planning method for drone inspection according to any one of claims 1 to 9, comprising: A map acquisition module is used to acquire a three-dimensional inspection map of the target inspection area, wherein the three-dimensional inspection map includes inspection equipment and obstacles; A function building module is used to build an objective function based on the inspection coverage, inspection energy consumption and inspection efficiency of the UAV in the target inspection area; The constraint building module is used to build constraints based on the inspection speed and inspection time of the drone and the shooting restrictions of the drone; The path generation module is used to obtain the initial planning path based on the objective function and constraints by adopting a local optimization strategy; The path optimization module is used to optimize the initial planned path using a global optimization strategy to complete the path planning for drone inspections.

Images