Double-layer optimized unmanned aerial vehicle path planning method and related equipment

Abstract

The invention relates to the technical field of unmanned aerial vehicle path planning, and provides a double-layer optimized unmanned aerial vehicle path planning method and related equipment. The method comprises the following steps: acquiring a base station position of a target unmanned aerial vehicle and positions of a plurality of inspection target points; constructing a path optimization function of the target unmanned aerial vehicle based on the position of the base station and the positions of all the inspection target points; solving the path optimization function by using a genetic algorithm and a tabu search algorithm to obtain an optimal path planning scheme of the target unmanned aerial vehicle; and controlling the target unmanned aerial vehicle to inspect all the inspection target points according to the optimal path planning scheme. The method provided by the invention can improve the precision of unmanned aerial vehicle path planning.

Description

Technical Field

[0001] The present application relates to the technical field of unmanned aerial vehicle path planning, and in particular to a double-layer optimized unmanned aerial vehicle path planning method and related equipment. Background Art

[0002] With the rapid development of wireless communication, sensor technology and UAV (Unmanned Aerial Vehicle) systems, UAVs are playing an increasingly important role in various applications, especially in the field of inspection. With their unique mobility and flexibility, UAVs have become an indispensable and important tool, playing an irreplaceable role in key tasks such as real-time monitoring, data collection and emergency response. When actually performing these tasks, in order to efficiently complete image acquisition, UAVs must carefully design flight paths based on diverse inspection needs, and the scientificity and rationality of path planning are directly related to the quality and efficiency of the entire inspection task.

[0003] In recent years, the academic community has shown a booming trend in the research of UAV path planning and task scheduling. Many scholars have devoted themselves to it and achieved a series of remarkable results. In the existing research, a considerable part of the literature focuses on the path planning problem of a single task. For example, some are committed to exploring the shortest path to reduce the flight distance, some focus on energy optimization to extend the endurance of the UAV, and some pursue the minimization of flight time to improve the timeliness of task execution. These studies provide valuable ideas and methods for UAV path planning from different perspectives, but there are still certain limitations when dealing with complex and changeable practical application scenarios.

[0004] Aiming at the core problem of UAV path planning, many algorithms have been proposed and widely used in various scenarios. Common path planning algorithms include classic heuristic search algorithms, such as variable step length in UAV path planning. Algorithms, classic graph algorithms, Dijkstra algorithms, etc. These algorithms perform well in simple scenarios, but when dealing with complex tasks and path planning problems under multiple constraints, they often have problems of large computational complexity and low solution efficiency. To overcome this shortcoming, researchers have begun to use evolutionary algorithms and swarm intelligence algorithms to solve UAV path planning problems. Genetic algorithms, as a classic optimization method, are widely used in UAV path planning, especially in multi-objective optimization and complex constraint problems. Genetic algorithms have global search capabilities and can effectively avoid local optimal solutions, but they are prone to high computational complexity in large-scale problems. In addition, tabu search (TS, Tabu Search) as a local search method is often used in optimization problems, especially for large-scale combinatorial optimization problems. Taboo search can break through the limitations of local optimality by continuously adjusting the neighborhood structure of the solution, thereby searching for the global optimal solution. In recent years, the combination of tabu search and genetic algorithms has become a hot research direction. Combining the advantages of both can effectively improve the efficiency and accuracy of path planning. In addition, nature-inspired algorithms such as Particle Swarm Optimization (PSO) and Ant Colony Optimization (ACO) have also been used in UAV path planning. These algorithms can find effective paths in complex environments by simulating the behavior of organisms in nature and have strong robustness.

[0005] In actual inspection tasks, the requirements of the task objectives for image acquisition accuracy are not single, but show obvious hierarchical differences, that is, there are high-value targets and low-value targets, and different types of targets have different requirements for image acquisition accuracy. According to the logical judgment of the actual situation, high-precision image acquisition data can meet the requirements of low-precision acquisition, and vice versa. This inconsistency in the image acquisition accuracy requirements of the task objectives undoubtedly greatly increases the complexity and challenge of path planning, resulting in low accuracy of drone path planning. Summary of the invention

[0006] The present application provides a dual-layer optimized UAV path planning method and related equipment, which can solve the problem of low accuracy in UAV path planning.

[0007] In a first aspect, an embodiment of the present application provides a dual-layer optimized UAV path planning method, the UAV path planning method comprising: Obtaining the base station location of the target UAV and the locations of multiple inspection target points; the inspection target point is one of a high-precision inspection target point or a low-precision inspection target point; A path optimization function for the target UAV is constructed based on the location of the base station and the locations of all inspection target points; the path optimization function is used to describe the total inspection time, total battery replacement time, and total flight time required for the target UAV to inspect all inspection target points; The path optimization function is solved by using genetic algorithm and taboo search algorithm to obtain the optimal path planning scheme of the target UAV. The individuals in the genetic algorithm are the path planning schemes of the target UAV, which include the inspection path and inspection accuracy of each flight of the target UAV. According to the optimal path planning scheme, the target UAV is controlled to inspect all inspection target points.

[0008] Optionally, the inspection target point is a high-precision inspection target point or a low-precision inspection target point, and the path optimization function is: ; in, Indicates the final total time it takes for the target drone to complete the inspection and return to the base station. Indicates the total flight time for the target UAV to complete all flight paths, Indicates the total inspection time of the target drone for all inspection target points. Indicates the total battery replacement time of the target drone at the base station: ; ; ; ; ; ; in, Indicates the total number of drone flights, Indicates the total number of inspection target points. Indicates The target drone flies from the base station to the Inspection target points, Indicates The target drone of the flight The inspection target points are returned to the base station. At that time, The inspection target point is the base station. Indicates The position of the first inspection target point is The distance between the locations of the inspection target points, Indicates the flight speed of the target drone, Indicates The first The inspection accuracy of each inspection target point, Indicates low precision, Indicates high precision, Indicates The first The inspection accuracy of each inspection target point, Indicates the time spent on high-precision inspection. Indicates the time spent on low-precision inspection. Indicates the time it takes to replace the battery. Indicates Does the voyage require battery replacement? Indicates that the battery needs to be replaced. Indicates that there is no need to replace the battery. Represents the set of inspection target points for actual low-precision inspection. Represents the set of inspection target points for actual high-precision inspection. Represents a set of low-precision inspection target points. Represents a set of high-precision inspection target points. Indicates the maximum flight time of the target drone in a single voyage. Represents a constraint.

[0009] Optionally, a genetic algorithm and a taboo search algorithm are used to solve the path optimization function to obtain an optimal path planning solution for the target UAV, including: Generate multiple individuals in genetic algorithms; The number of iterations is increased by 1 to determine whether the number of iterations is greater than or equal to the maximum number of iterations; If so, the best individual is selected from all individuals according to the path optimization function, and the path planning scheme described by the best individual is used as the optimal path planning scheme; Otherwise, use the genetic algorithm to update all individuals to obtain multiple updated individuals, and use the taboo search algorithm to optimize the search for each updated individual to obtain the better individual corresponding to each updated individual, and use the multiple better individuals as multiple individuals, return the number of iterations plus 1, and determine whether the number of iterations is greater than or equal to the maximum number of iterations.

[0010] Optionally, the best individual is selected from all individuals according to the path optimization function, including: For each individual, substitute the path planning scheme described by the individual into the path optimization function and calculate the final total time corresponding to the individual; The individual corresponding to the final total time with the smallest value is regarded as the optimal individual.

[0011] Optionally, all individuals are updated using a genetic algorithm to obtain multiple updated individuals, including: Calculate the fitness of each individual according to the path optimization function; Select multiple parent individuals from all individuals according to all fitness; Perform a crossover operation on all parent individuals to obtain multiple offspring individuals, and perform a mutation operation on all offspring individuals to obtain multiple mutant offspring individuals; The fitness of each mutant offspring individual is calculated according to the path optimization function, and all mutant offspring individuals are sorted from large to small according to their fitness, and the first multiple mutant offspring individuals in the sorting result are selected as update individuals.

[0012] Optionally, a taboo search algorithm is used to optimize the search for each update individual to obtain a better individual corresponding to each update individual, including: For each update individual, perform the following steps: Determine whether there is a voyage that mixes low-precision inspection target points and high-precision inspection target points in the path planning scheme for updating the individual description; If yes, use the tabu search algorithm to search the neighborhood of the updated individual to obtain the better individual corresponding to the updated individual; Otherwise, the updated individual is regarded as the better individual corresponding to itself.

[0013] Optionally, a tabu search algorithm is used to perform a neighborhood search on the update individual to obtain a better individual corresponding to the update individual, including: Initialize the taboo table; Selecting a target neighborhood search operation from a plurality of neighborhood search operations; Performing a neighborhood search on the update individual using a target neighborhood search operation to obtain a candidate individual of the update individual, and adding the target neighborhood search operation to the taboo table; When the taboo list does not include all neighborhood search operations, a target neighborhood search operation is selected from all other neighborhood search operations that have not been added to the taboo list, and the target neighborhood search operation is used to perform a neighborhood search on the update individual to obtain a candidate individual of the update individual, and the target neighborhood search operation is added to the taboo list; When all neighborhood search operations are included in the taboo table, the fitness of each candidate individual is calculated according to the path optimization function, and the candidate individual with the largest fitness value is used as the better individual for updating the individual.

[0014] Optionally, multiple neighborhood search operations, including: First neighborhood search operation: in a voyage with mixed low-precision inspection target points and high-precision inspection target points, all high-precision patrol target points are deleted from the voyage; Second neighborhood search operation: in a voyage where low-precision inspection target points and high-precision inspection target points are mixed, all low-precision patrol target points are deleted from the voyage; The third neighborhood search operation: perform neighborhood transformation on the voyage of the mixed low-precision inspection target point and the high-precision inspection target point to obtain another voyage of the mixed low-precision inspection target point and the high-precision inspection target point.

[0015] In a second aspect, an embodiment of the present application provides a dual-layer optimized drone path planning device, comprising: An acquisition module is used to acquire the base station position of the target UAV and the positions of multiple inspection target points; the inspection target point is one of a high-precision inspection target point or a low-precision inspection target point; A construction module is used to construct a path optimization function of the target UAV based on the location of the base station and the locations of all inspection target points; the path optimization function is used to describe the total inspection time, total battery replacement time and total flight time required for the target UAV to inspect all inspection target points; The solution module is used to solve the path optimization function using the genetic algorithm and the taboo search algorithm to obtain the optimal path planning scheme of the target UAV; the individual in the genetic algorithm is the path planning scheme of the target UAV, and the path planning scheme includes the inspection path of each voyage of the target UAV and the inspection accuracy; The inspection module is used to control the target UAV to inspect all inspection target points according to the optimal path planning solution.

[0016] In a third aspect, an embodiment of the present application provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned double-layer optimized drone path planning method when executing the above-mentioned computer program.

[0017] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, it implements the above-mentioned double-layer optimized drone path planning method.

[0018] The above solution of the present application has the following beneficial effects: In some embodiments of the present application, the base station location of the target UAV and the locations of multiple inspection target points are obtained, and then the path optimization function of the target UAV is constructed based on the base station location and the locations of all inspection target points, and then the path optimization function is solved using the genetic algorithm and the taboo search algorithm to obtain the optimal path planning scheme of the target UAV, and finally the target UAV is controlled to inspect all inspection target points according to the optimal path planning scheme. Among them, constructing the path optimization function can represent the final total time required for the inspection, and combining the genetic algorithm and the taboo search algorithm can effectively overcome the limitations of the two algorithms, improve the global search capability and local optimization performance of the path planning, and have good adaptability. By solving the path optimization function in this way, the final total time required for the inspection can be considered, the accuracy and execution efficiency of the path planning scheme can be improved, and the accuracy of the UAV path planning can be effectively improved.

[0019] Other beneficial effects of the present application will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 A flowchart of a double-layer optimized drone path planning method provided in one embodiment of the present application; Figure 2 A schematic diagram of an individual in a genetic algorithm provided in an embodiment of the present application; Figure 3 A schematic diagram of a sequential crossover operation provided by an embodiment of the present application; Figure 4 A schematic diagram of the execution of the 3-Opt algorithm provided in one embodiment of the present application; Figure 5 A schematic diagram of the distribution of multiple inspection target points provided in an embodiment of the present application; Figure 6 A schematic diagram of an iterative convergence curve provided in an embodiment of the present application; Figure 7 A schematic diagram of an optimal path planning scheme of GA-VRP provided in an embodiment of the present application; Figure 8 A schematic diagram of an optimal path planning solution for GA-VRPHD provided in an embodiment of the present application; Fig. 9A schematic diagram of an optimal path planning scheme of the method of the present application provided in one embodiment of the present application; Fig.10 A schematic diagram of the structure of a double-layer optimized drone path planning device provided in one embodiment of the present application; Fig.11 A schematic diagram of the structure of a terminal device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0022] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0023] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0024] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0025] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0026] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0027] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0028] In order to solve the problem of low accuracy of existing UAV path planning, the embodiment of the present application provides a double-layer optimized UAV path planning method, which obtains the base station location of the target UAV and the locations of multiple inspection target points, and then constructs the path optimization function of the target UAV based on the base station location and the locations of all inspection target points, and then solves the path optimization function using a genetic algorithm and a taboo search algorithm to obtain the optimal path planning scheme for the target UAV, and finally controls the target UAV to inspect all inspection target points according to the optimal path planning scheme. Among them, constructing a path optimization function can represent the final total time required for inspection, and combining the genetic algorithm and the taboo search algorithm can effectively overcome the limitations of the two algorithms, improve the global search capability and local optimization performance of path planning, and has good adaptability. By solving the path optimization function in this way, the final total time required for inspection can be considered, the accuracy and execution efficiency of the path planning scheme can be improved, and the accuracy of the UAV path planning can be effectively improved.

[0029] Next, an exemplary description is given of the double-layer optimized UAV path planning method provided in this application.

[0030] like Figure 1 As shown, the dual-layer optimized UAV path planning method provided in this application includes the following steps: Step 11, obtain the base station location of the target drone and the locations of multiple inspection target points.

[0031] The inspection target point is a target that needs to be inspected by a drone. For example, the inspection target point may be a railway track that needs to be inspected.

[0032] The inspection target points are high-precision inspection target points or low-precision inspection target points. The multiple inspection target points may include high-precision inspection target points and low-precision inspection target points, or may include only high-precision inspection target points, or may include only low-precision inspection target points. The inspection accuracy required for high-precision inspection target points is higher than that for low-precision inspection target points. For example, when using drones to collect images of inspection target points, the accuracy of images of high-precision inspection target points is higher than that of images of low-precision inspection target points.

[0033] In some embodiments of the present application, the base station location of the target drone and the locations of multiple inspection target points can be obtained through a global positioning system or the like.

[0034] It should be noted that the accuracy of drone inspections corresponds to the flight altitude of the drone (for example, when collecting high-precision images, the flight altitude of the drone is lower than when collecting low-precision images). Therefore, in a drone inspection voyage, the drone can only work with the same inspection accuracy for all targets, and according to actual needs, high-precision inspections can meet the requirements of low-precision inspection target points, but low-precision inspections cannot meet the requirements of high-precision inspection target points, and the time required for drones to inspect high-precision inspection target points is greater than the time required for drones to inspect low-precision inspection target points.

[0035] Step 12: construct a path optimization function for the target UAV based on the base station location and the locations of all inspection target points.

[0036] The above path optimization function is used to describe the total inspection time, total battery replacement time and total flight time required for the target UAV to inspect all inspection target points.

[0037] The total inspection time refers to the total time spent by the target UAV to inspect each inspection target point after it arrives at the inspection target point.

[0038] The total battery replacement time refers to the total time spent replacing batteries during the inspection of all inspection target points by the target drone.

[0039] The total flight time refers to the total time spent by the target UAV in the process of inspecting all inspection target points.

[0040] The above path optimization function is: ; in, Indicates the final total time it takes for the target drone to complete the inspection and return to the base station. Indicates the total flight time for the target UAV to complete all flight paths, Indicates the total inspection time of the target drone for all inspection target points. Indicates the total battery replacement time of the target drone at the base station: ; ; ; ; ; ; in, Indicates the total number of drone flights, Indicates the total number of inspection target points. Indicates The target drone flies from the base station to the Inspection target points, Indicates The target drone of the flight The inspection target points are returned to the base station. This means that all flights must depart from and land at the base station. At that time, The inspection target point is the base station. Indicates The position of the first inspection target point is The distance between the locations of the inspection target points, Indicates the flight speed of the target drone, Indicates The first The inspection accuracy of each inspection target point, Indicates low precision, Indicates high precision, Indicates The first The inspection accuracy of each inspection target point, Indicates the time spent on high-precision inspection. Indicates the time spent on low-precision inspection. Indicates the time it takes to replace the battery. Indicates Does the voyage require battery replacement? Indicates that the battery needs to be replaced. Indicates that there is no need to replace the battery. Represents the set of inspection target points for actual low-precision inspection. Represents the set of inspection target points for actual high-precision inspection. Represents a set of low-precision inspection target points. Represents a set of high-precision inspection target points. Indicates the maximum flight time of the target drone in a single voyage. Represents a constraint.

[0041] It should be noted that the distance between the positions of two inspection target points can be the Euclidean distance, and the expression of the Euclidean distance is: ; in, Indicates The position of the first inspection target point is The Euclidean distance between the locations of the inspection target points, Indicates The horizontal coordinate position of the inspection target point, Indicates The horizontal coordinate position of the inspection target point, Indicates The vertical coordinate position of the inspection target point, Indicates The vertical coordinate position of the inspection target point.

[0042] Step 13, using genetic algorithm and taboo search algorithm to solve the path optimization function and obtain the optimal path planning solution for the target UAV.

[0043] The individuals in the above genetic algorithm are the path planning schemes of the target UAV, and the path planning schemes include the inspection path and inspection accuracy of each voyage of the target UAV.

[0044] In some embodiments of the present application, the step of solving the path optimization function using the genetic algorithm and the taboo search algorithm to obtain the optimal path planning solution for the target UAV includes: The first step is to generate multiple individuals in the genetic algorithm.

[0045] Specifically, the above individuals include three coding structures: 1. Target point access sequence encoding The target point visit sequence code represents the visit sequence of the cruise targets. This coding structure can directly represent the sequence of the cruise targets.

[0046] 2. Voyage classification code It is used to divide the target point visit sequence, determine which cruise targets should be assigned to a certain voyage, and ensure that the cruise targets of each voyage do not exceed the battery capacity limit of the drone. This step is a key link in drone mission planning, aiming to optimize the drone's flight path and task allocation to improve efficiency and avoid energy depletion.

[0047] 3. Precision allocation encoding This code determines the accuracy used for image acquisition for each voyage obtained by the voyage segmentation code. The voyage accuracy code is directly related to the target image acquisition time in the total time. In this code, the division of high-precision voyages and low-precision voyages can be distinguished by setting different marks, 0 for low accuracy and 1 for high accuracy.

[0048] By combining the above three coding structures to obtain the individuals in the genetic algorithm, the UAV inspection path problem can be converted into an individual representation suitable for solving by the genetic algorithm.

[0049] For example, individuals and their decoding are as follows Figure 2 As shown, the individuals Figure 2 As shown in a, it includes target point access sequence code, voyage division code and precision allocation code. The target point access sequence code is used to describe the access sequence of all inspection target points. The voyage division code is used to describe the voyage division of the access sequence of all inspection target points (after dividing the target point access sequence code according to the voyage division code and inserting the base station code, the number of voyages and the inspection path of each voyage are obtained). The precision allocation code is used to describe the inspection accuracy of each voyage. The target point access sequence code 146235 indicates that the inspection sequence of the target UAV should be based on the codes of the 6 inspection target points, and the inspection should be carried out in the order of 146235. The voyage division code 24 indicates that the voyage is divided at the second inspection target point and the fourth inspection target point, that is, the second inspection target point and the fourth inspection target point are respectively used as the last inspection target point of the corresponding voyage. The precision allocation code 001 indicates that the inspection accuracy of the three voyages obtained according to the voyage division code is low accuracy, low accuracy and high accuracy respectively. The process of decoding this individual is as follows Figure 2 As shown in Figure 2, the target point access sequence code is divided according to the voyage division code, and the division results are: 14, 62, 35. The base station number 0 is inserted at the head and tail, and the precision allocation code is read to obtain three voyages, namely, low-precision voyage 1: 0140 (indicating that the inspection path of this voyage is base station-cruise target point 1-cruise target point 4-base station), low-precision voyage 2: 0620 (indicating that the inspection path of this voyage is base station-cruise target point 6-cruise target point 2-base station), and high-precision voyage 1: 0350 (indicating that the inspection path of this voyage is base station-cruise target point 3-cruise target point 5-base station).

[0050] The second step is to add 1 to the number of iterations and determine whether the number of iterations is greater than or equal to the maximum number of iterations.

[0051] The initial value of the above iteration number is 0.

[0052] If so, the best individual is selected from all individuals according to the path optimization function, and the path planning scheme described by the best individual is used as the optimal path planning scheme.

[0053] Specifically, for each individual, the path planning scheme described by the individual is substituted into the path optimization function to calculate the final total time corresponding to the individual; the individual corresponding to the final total time with the smallest value is taken as the optimal individual.

[0054] Otherwise, use the genetic algorithm to update all individuals to obtain multiple updated individuals, and use the taboo search algorithm to optimize the search for each updated individual to obtain the better individual corresponding to each updated individual, and use the multiple better individuals as multiple individuals, return the number of iterations plus 1, and determine whether the number of iterations is greater than or equal to the maximum number of iterations.

[0055] It should be noted that the above step of using the genetic algorithm to update all individuals to obtain multiple updated individuals includes: First, the fitness of each individual is calculated according to the path optimization function.

[0056] Specifically, the path planning scheme described by each individual is substituted into the path optimization function, the final total time corresponding to the individual is calculated, and the final total time is used as the fitness of the corresponding individual.

[0057] Then, multiple parent individuals are selected from all individuals according to all fitness.

[0058] Exemplarily, a selection operation in a genetic algorithm (such as ranking selection, roulette wheel selection, etc.) may be used to select multiple parent individuals from all individuals according to all fitness values.

[0059] Then, a crossover operation is performed on all parent individuals to obtain multiple offspring individuals, and a mutation operation is performed on all offspring individuals to obtain multiple mutant offspring individuals.

[0060] Specifically, a sequential crossover operation can be used to perform a crossover operation on all parent individuals to obtain multiple offspring individuals. A mutation operation can be performed on each offspring individual using site mutation, exchange mutation, Gaussian mutation, etc. to obtain a mutated offspring individual corresponding to each offspring individual.

[0061] For example, Figure 3As shown, based on the target point access sequence encoding in two parent individuals (parent 1 and parent 2), the target point access sequence encoding in child 1 is obtained by sequential crossover operation. First, a starting point 2 and an end point 4 are randomly selected from parent 1. The sequence segment from the starting point to the end point is extracted to generate the basis of child 1, and the nodes in parent 2 that are repeated with the selected sequence are deleted, and the remaining nodes in parent 2 are sequentially filled into the blanks of child 1 to generate a complete code for child 1. Child 2 can be generated in the same way.

[0062] Finally, the fitness of each mutant offspring individual is calculated according to the path optimization function, and all mutant offspring individuals are sorted from large to small according to their fitness, and the first multiple mutant offspring individuals in the sorting results are selected as update individuals.

[0063] Specifically, the path planning scheme described by each mutant offspring individual is substituted into the path optimization function, the final total time corresponding to the mutant offspring individual is calculated, and the final total time is used as the fitness of the corresponding mutant offspring individual.

[0064] It should be noted that the above step of using the taboo search algorithm to optimize the search for each update individual to obtain the better individual corresponding to each update individual includes: For each update individual, perform the following steps: Determine whether there is a voyage with mixed low-precision inspection target points and high-precision inspection target points in the path planning scheme for updating the individual description.

[0065] If so, the taboo search algorithm is used to perform neighborhood search on the updated individual to obtain the better individual corresponding to the updated individual.

[0066] Specifically, the taboo table is initialized, and then a target neighborhood search operation is selected from multiple neighborhood search operations (the target neighborhood search operation can be randomly selected), and then the target neighborhood search operation is used to perform a neighborhood search on the update individual to obtain a candidate individual of the update individual, and the target neighborhood search operation is added to the taboo table; When the taboo list does not include all neighborhood search operations (i.e., the multiple neighborhood search operations described above), a target neighborhood search operation is selected from all other neighborhood search operations that have not been added to the taboo list (the target neighborhood search operation can be randomly selected), and the step of performing a neighborhood search on the update individual using the target neighborhood search operation to obtain a candidate individual of the update individual and adding the target neighborhood search operation to the taboo list is returned; When all neighborhood search operations are included in the taboo table, the fitness of each candidate individual is calculated according to the path optimization function (the path planning scheme described by the candidate individual is substituted into the path optimization function to obtain the final total time corresponding to the candidate individual, and the final total time is used as the fitness of the candidate individual), and the candidate individual with the largest fitness value is used as the better individual for updating the individual.

[0067] Otherwise (that is, in the path planning scheme described by the updated individual, there is no voyage that mixes low-precision inspection target points and high-precision inspection target points), the updated individual will be used as the corresponding better individual.

[0068] It should be noted that the above multiple neighborhood search operations include: First neighborhood search operation: in a voyage with mixed low-precision inspection target points and high-precision inspection target points, all high-precision patrol target points are deleted from the voyage; Second neighborhood search operation: in a voyage where low-precision inspection target points and high-precision inspection target points are mixed, all low-precision patrol target points are deleted from the voyage; The third neighborhood search operation: perform neighborhood transformation on the voyage of mixed low-precision inspection target points and high-precision inspection target points to obtain another voyage of mixed low-precision inspection target points and high-precision inspection target points. That is, by adjusting the access order of each inspection target point within the voyage to optimize the overall efficiency or cost, this can be achieved through the 3-Opt heuristic algorithm (3-Opt, 3-Opt Heuristic Algorithm).

[0069] Exemplarily, the 3-Opt algorithm is performed as follows Figure 4 As shown in Figure 4, for the six inspection target points ABCDEF, starting from the initial roadbed diagram shown in Figure 4 (a), three edges are randomly deleted, and then the remaining edges are randomly combined. For a case where three edges have been deleted, there are 8 possible solutions, such as Figure 4 (b) to Figure 4 As shown in (i), the dotted line is a random edge generated after deleting the edge. Figure 4 (b) is the first possible cruise path, Figure 4 (c) is the second possible cruise path, Figure 4 (d) is the third possible cruise path, Figure 4 (e) is the fourth possible cruise path, Figure 4 (f) is the fifth possible cruise path, Figure 4 (g) is the sixth possible cruise path, Figure 4 (h) is the seventh possible cruise path, Figure 4 (i) is the eighth possible cruise path.

[0070] Step 14: Control the target UAV to inspect all inspection target points according to the optimal path planning solution.

[0071] Specifically, the control system of the target UAV can be used to enable the UAV to execute each voyage in sequence according to the inspection path and inspection accuracy of each voyage in the optimal path planning plan, and complete the inspection of all inspection target points.

[0072] It is worth mentioning that constructing a path optimization function can represent the total time required for inspection. Combining the genetic algorithm and the taboo search algorithm can effectively overcome the limitations of the two algorithms, improve the global search capability and local optimization performance of path planning, and have good adaptability. By solving the path optimization function in this way, the total time required for inspection can be taken into account, the accuracy and execution efficiency of the path planning scheme can be improved, and the accuracy of UAV path planning can be effectively improved.

[0073] The method of the present application is illustrated below with reference to a specific example.

[0074] Matlab2024b simulation software is used to simulate and compare the algorithm of this application and other basic algorithms. The simulation area is 100 km 100 km, where the drone take-off base station is located at (50, 50), and the distribution of 20 target points to be imaged (i.e., inspection target points) is shown in Figure 5 , of which 4 target points are high-precision acquisition targets ( Figure 5 pentagonal point), 16 target points are low-precision acquisition targets ( Figure 5 middle square point), Figure 5 The points of the five-pointed star are base stations, and the horizontal and vertical axes are coordinate axes. The environmental simulation parameters for UAV path planning are shown in Table 1, and the hyperparameters of the genetic algorithm and taboo search are shown in Table 2.

[0075] Table 1 ; Table 2 ; There are two comparative algorithms for experimental simulation: one is to directly divide the target into two tasks, a high-precision vehicle routing problem (VRP) and a low-precision VRP, according to the initial image acquisition accuracy requirements. This algorithm also uses a genetic algorithm to plan two independent VRP problems. The algorithm is called the Genetic Algorithm for Vehicle Routing Problem (GA-VRP); the other is to model the problem as a drone path planning involving hierarchical optimization, and use a genetic algorithm that does not include taboo search to solve it, which is the Genetic Algorithm for Vehicle Routing Problem with Heterogeneous Demands (GA-VRPHD).

[0076] The iterative convergence curves and local magnification diagrams of the three algorithms are shown in Figure 2. Figure 6 shown. Figure 6 The horizontal axis represents the iteration rounds, and the vertical axis represents the shortest time in hours. The curve with dots represents the iterative convergence curve of GA-VRP, the curve with square dots represents the iterative convergence curve of GA-VRPHD, and the curve with diamonds represents the iterative convergence curve of the method of the present application (GA-TS). 16.2766 is the total time required for the final stable UAV cycle of the method of the present application.

[0077] It can be seen from the iterative convergence curve that GA-TS achieved the best result, taking 16.2766 hours, which is better than GA-VRP and GA-VRPHD. Moreover, by observing the local zoom curve, it can be found that GA-TS also has the fastest convergence speed. This is in line with expectations. In the two-layer optimization, the use of the taboo search algorithm can better divide the low-precision and high-precision voyages, so as to better escape the local optimal solution. By comparing GA-VRP and GA-VRPHD, it can be found that the results of GA-VRPHD are better than those of GA-VRP, indicating that the two-layer optimization modeling of the UAV is better than directly and rigidly dividing the task into high-precision collection voyages and low-precision collection voyages, and then performing path planning separately.

[0078] The optimal path planning solution obtained by GA-VRP is as follows: Figure 7 As shown, the horizontal and vertical axes are coordinate axes, the points in the pentagon represent high-precision inspection target points, the points in the quadrilateral represent low-precision inspection target points, the points in the five-pointed star represent base stations, the solid lines represent low-precision voyages, and the dotted lines represent high-precision voyages.

[0079] From the results, we can see that the genetic algorithm works independently to achieve high-precision and low-precision UAV path planning. In the obtained path, there is no situation where low-precision targets and high-precision targets are mixed in one voyage. This rigid voyage division ignores some high-precision and low-precision combination voyages. This combination of voyages with short distances but better overall results is searched. Because it is only a simple VRP modeling, the solution complexity is low, which is related to Figure 6 The iterative convergence curve in is consistent, and the optimal solution is obtained after about 100 iterations.

[0080] The optimal path planning solution obtained by GA-VRPHD is as follows: Figure 8 As shown, the horizontal and vertical axes are coordinate axes, the points in the pentagon represent high-precision inspection target points, the points in the quadrilateral represent low-precision inspection target points, the points in the five-pointed star represent base stations, the solid lines represent low-precision voyages, and the dotted lines represent high-precision voyages.

[0081] It can be seen that in the path of high-precision voyage 1, high-precision voyage 2, and high-precision voyage 3, there are high-precision targets and low-precision targets mixed in one voyage, but there is only one high-precision target in high-precision voyage 1 and high-precision voyage 2. Due to the collection equipment limitation of each collection voyage, only one collection accuracy can be used, so other low-precision collection targets in high-precision voyage 1 and high-precision voyage 2 use high-precision equipment to collect data, which takes longer to collect. Because GA-VRPHD does not perform task optimization for low-level voyage division, but only confirms the matching of low-precision and high-precision tasks, GA-VRPH is prone to fall into the local optimum for this two-layer optimization algorithm. Figure 6 It can be seen that a feasible solution was obtained after 243 rounds of iteration.

[0082] The optimal path planning solution obtained by the method of this application is as follows Fig. 9 As shown, the horizontal and vertical axes are coordinate axes, the points in the pentagon represent high-precision inspection target points, the points in the quadrilateral represent low-precision inspection target points, the points in the five-pointed star represent base stations, the solid lines represent low-precision voyages, and the dotted lines represent high-precision voyages.

[0083] exist Fig. 9 There is a path intersection between low-precision voyage 3 and high-precision voyage 1. In traditional path planning algorithms, it is generally believed that the optimal result has not been found. However, because the time required for target image acquisition for high-precision voyages is longer than that for low-precision voyages, the path planning obtained is actually better. There is no intersection between low-precision voyages, and there is no intersection between high-precision voyages. Due to the hierarchical nature of two-layer optimization and the conflict between low-level optimization and high-level optimization, it is almost impossible for two-layer optimization to find the optimal solution. GA-TS Figure 6In the iterative convergence curve of the three algorithms, it converges the fastest and obtains the best results.

[0084] It can be seen that the method of the present application can effectively solve the problem of reasonable allocation of high-precision and low-precision inspections, and fully consider the battery capacity limitation in path planning, thereby optimizing the inspection efficiency of drones. Compared with the traditional single algorithm, the method of the present application has significant advantages in the collaborative optimization of task precision division and path planning, and can improve the accuracy and execution efficiency of path planning in complex inspection tasks.

[0085] In addition, the research of this application also shows that the hybrid optimization method combining taboo search and genetic algorithm can effectively overcome the limitations of the two algorithms and improve the global search capability and local optimization performance of the path planning algorithm. Especially when facing practical problems such as differences in mission target accuracy and battery constraints, this method has good adaptability and problem-solving capabilities. In addition, with the continuous advancement of drone technology and computing power, future research can consider path planning problems in large-scale, multi-task scenarios to further improve the application effect of drones in fields such as inspection.

[0086] The following is an exemplary description of the double-layer optimized UAV path planning device provided in this application.

[0087] like Fig.10 As shown, the embodiment of the present application provides a dual-layer optimized UAV path planning device, and the dual-layer optimized UAV path planning device 1000 includes: The acquisition module 1001 is used to acquire the base station position of the target UAV and the positions of multiple inspection target points; the inspection target point is one of a high-precision inspection target point or a low-precision inspection target point; Construction module 1002, used to construct a path optimization function of the target UAV based on the base station location and the locations of all inspection target points; the path optimization function is used to describe the total inspection time, total battery replacement time, and total flight time required for the target UAV to inspect all inspection target points; The solution module 1003 is used to solve the path optimization function by using a genetic algorithm and a taboo search algorithm to obtain an optimal path planning scheme for the target UAV; the individuals in the genetic algorithm are the path planning schemes for the target UAV, and the path planning schemes include the inspection path and inspection accuracy of each voyage of the target UAV; The inspection module 1004 is used to control the target UAV to inspect all inspection target points according to the optimal path planning solution.

[0088] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0089] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0090] like Fig.11 As shown, an embodiment of the present application provides a terminal device. The terminal device D10 of this embodiment includes: at least one processor D100 ( Fig.11 Only one processor is shown in the figure), a memory D101, and a computer program D102 stored in the memory D101 and executable on the at least one processor D100, wherein the processor D100 implements the steps of any of the above-mentioned method embodiments when executing the computer program D102.

[0091] Specifically, when the processor D100 executes the computer program D102, it obtains the base station location of the target drone and the locations of multiple inspection target points, then constructs the path optimization function of the target drone based on the base station location and the locations of all inspection target points, and then uses the genetic algorithm and the taboo search algorithm to solve the path optimization function to obtain the optimal path planning scheme of the target drone, and finally controls the target drone to inspect all inspection target points according to the optimal path planning scheme. Among them, constructing the path optimization function can represent the final total time required for inspection, and combining the genetic algorithm and the taboo search algorithm can effectively overcome the limitations of the two algorithms, improve the global search capability and local optimization performance of path planning, and have good adaptability. By solving the path optimization function in this way, the final total time required for inspection can be considered, the accuracy and execution efficiency of the path planning scheme can be improved, and the accuracy of drone path planning can be effectively improved.

[0092] The processor D100 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0093] In some embodiments, the memory D101 may be an internal storage unit of the terminal device D10, such as a hard disk or memory of the terminal device D10. In other embodiments, the memory D101 may also be an external storage device of the terminal device D10, such as a plug-in hard disk, a smart memory card (SMC, SmartMedia Card), a secure digital (SD, Secure Digital) card, a flash card (Flash Card), etc. equipped on the terminal device D10. Further, the memory D101 may also include both an internal storage unit of the terminal device D10 and an external storage device. The memory D101 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as the program code of the computer program, etc. The memory D101 may also be used to temporarily store data that has been output or is to be output.

[0094] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0095] An embodiment of the present application provides a computer program product. When the computer program product runs on a terminal device, the terminal device can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0096] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the dual-layer optimized drone path planning method device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0097] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0098] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0099] The above is a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles described in the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A two-layer optimized UAV path planning method, characterized in that: include: Obtain the base station location of the target drone and the locations of multiple inspection target points; Constructing a path optimization function for the target UAV based on the location of the base station and the locations of all inspection target points; The path optimization function is used to describe the total inspection time, total battery replacement time and total flight time required for the target UAV to inspect all inspection target points; The path optimization function is solved by using a genetic algorithm and a taboo search algorithm to obtain an optimal path planning scheme for the target UAV; the individuals in the genetic algorithm are the path planning schemes for the target UAV, and the path planning schemes include the inspection path and inspection accuracy of each voyage of the target UAV; The target UAV is controlled to inspect all inspection target points according to the optimal path planning solution.

2. The UAV path planning method according to claim 1, characterized in that: The inspection target point is a high-precision inspection target point or a low-precision inspection target point; The path optimization function is: ； in, Indicates the final total time it takes for the target drone to complete the inspection and return to the base station. represents the total flight time of the target UAV to complete all flight paths, Indicates the total inspection time of the target UAV for inspecting all inspection target points. Indicates the total battery replacement time of the target drone at the base station: ； ； ； ； ； ； in, Indicates the total number of drone flights, Indicates the total number of inspection target points. Indicates Whether the target drone in the flight Fly to the inspection target point Inspection target points, Indicates The target drone flies from the base station to the Inspection target points, Indicates The target drone of the flight The inspection target points are returned to the base station. At that time, The inspection target point is the base station. Indicates The position of the first inspection target point is The distance between the locations of the inspection target points, Indicates the flight speed of the target drone, Indicates The first The inspection accuracy of each inspection target point, Indicates low precision, Indicates high precision, Indicates The first The inspection accuracy of each inspection target point, Indicates the time spent on high-precision inspection. Indicates the time spent on low-precision inspection. Indicates the time it takes to replace the battery. Indicates Does the voyage require battery replacement? Indicates that the battery needs to be replaced. Indicates that there is no need to replace the battery. Represents the set of inspection target points for actual low-precision inspection. Represents the set of inspection target points for actual high-precision inspection. Represents a set of low-precision inspection target points. Represents a set of high-precision inspection target points. Indicates the maximum flight time of the target UAV in a single flight. Represents a constraint.

3. The UAV path planning method according to claim 2, characterized in that: The method of solving the path optimization function by using a genetic algorithm and a taboo search algorithm to obtain an optimal path planning solution for the target UAV includes: Generate multiple individuals in genetic algorithms; The number of iterations is increased by 1, and it is determined whether the number of iterations is greater than or equal to the maximum number of iterations; If yes, then according to the path optimization function, the best individual is selected from all individuals, and the path planning scheme described by the best individual is used as the optimal path planning scheme; Otherwise, all individuals are updated using a genetic algorithm to obtain multiple updated individuals, and each updated individual is optimized and searched using a taboo search algorithm to obtain a better individual corresponding to each updated individual, and multiple better individuals are taken as multiple individuals, and the number of iterations plus 1 is returned to determine whether the number of iterations is greater than or equal to the maximum number of iterations.

4. The UAV path planning method according to claim 3, characterized in that: The step of selecting the best individual from all individuals according to the path optimization function comprises: For each individual, the path planning scheme described by the individual is substituted into the path optimization function to calculate the final total time corresponding to the individual; The individual corresponding to the final total time with the smallest value is regarded as the optimal individual.

5. The UAV path planning method according to claim 3, characterized in that: The genetic algorithm is used to update all individuals to obtain multiple updated individuals, including: Calculate the fitness of each individual according to the path optimization function; Select multiple parent individuals from all individuals according to all fitness; Perform a crossover operation on all parent individuals to obtain multiple offspring individuals, and perform a mutation operation on all offspring individuals to obtain multiple mutant offspring individuals; The fitness of each mutant offspring individual is calculated according to the path optimization function, and all mutant offspring individuals are sorted from large to small according to their fitness, and the first multiple mutant offspring individuals in the sorting result are selected as update individuals.

6. The UAV path planning method according to claim 5, characterized in that: The method of using the taboo search algorithm to optimize and search each updated individual to obtain a better individual corresponding to each updated individual includes: For each update individual, perform the following steps: Determine whether there is a voyage that mixes low-precision inspection target points and high-precision inspection target points in the path planning scheme described by the updated individual; If yes, use the tabu search algorithm to perform neighborhood search on the updated individual to obtain a better individual corresponding to the updated individual; Otherwise, the updated individual is taken as the better individual corresponding to itself.

7. The UAV path planning method according to claim 6, characterized in that: The step of performing a neighborhood search on the updated individual using a taboo search algorithm to obtain a better individual corresponding to the updated individual includes: Initialize the taboo table; Selecting a target neighborhood search operation from a plurality of neighborhood search operations; Performing a neighborhood search on the update individual using a target neighborhood search operation to obtain a candidate individual of the update individual, and adding the target neighborhood search operation to a taboo table; When the taboo list does not include all neighborhood search operations, a target neighborhood search operation is selected from all other neighborhood search operations that have not been added to the taboo list, and the target neighborhood search operation is used to perform a neighborhood search on the update individual to obtain a candidate individual of the update individual, and the target neighborhood search operation is added to the taboo list; When the taboo table includes all neighborhood search operations, the fitness of each candidate individual is calculated according to the path optimization function, and the candidate individual with the largest fitness value is used as the better individual of the updated individual.

8. The UAV path planning method according to claim 7, characterized in that: The multiple neighborhood search operations include: First neighborhood search operation: in a voyage with mixed low-precision patrol target points and high-precision patrol target points, all high-precision patrol target points are deleted from the voyage; Second neighborhood search operation: in a voyage where low-precision patrol target points and high-precision patrol target points are mixed, all low-precision patrol target points are deleted from the voyage; The third neighborhood search operation: perform neighborhood transformation on the voyage of the mixed low-precision inspection target point and the high-precision inspection target point to obtain another voyage of the mixed low-precision inspection target point and the high-precision inspection target point.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the double-layer optimized drone path planning method as described in any one of claims 1 to 8 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the double-layer optimized UAV path planning method as described in any one of claims 1 to 8 is implemented.

Images