Unmanned aerial vehicle path planning method and device, electronic equipment and storage medium

Abstract

The invention is suitable for the technical field of unmanned aerial vehicle path planning, and provides an unmanned aerial vehicle path planning method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining path planning information, the path planning information comprises a starting point, a terminal point and obstacle information of an unmanned aerial vehicle; based on the path planning information, determining a feasible path by adopting a fast expansion random tree algorithm; converting the feasible path into a plurality of candidate paths represented by a chromosome form in a genetic algorithm; and based on each candidate path, determining an optimal path by adopting a genetic algorithm. According to the technical scheme provided by the invention, the rapid exploration capability of the rapid extension random tree algorithm and the global optimization capability of the genetic algorithm are combined, rapid generation and fine adjustment of the path are realized, the smoothness and the travelability of the planned path are enhanced, the finally obtained planned path is better, and the path planning efficiency is improved. The efficiency and accuracy of unmanned aerial vehicle path planning in a complex environment are improved.

Description

Technical Field

[0001] The present application belongs to the technical field of unmanned aerial vehicle path planning, and specifically relates to a unmanned aerial vehicle path planning method, a planning device, an electronic device and a storage medium. Background Art

[0002] With the continuous development of drone technology, drones are widely used in various fields. Path planning has a great impact on the actual flight of drones.

[0003] The current UAV path planning is usually implemented by a single path planning algorithm, such as the rapidly-exploring random tree (RRT) algorithm. However, a single path planning algorithm has limitations due to its algorithm characteristics, which makes the paths planned by the UAV in different environments poor, resulting in low accuracy of UAV path planning.

[0004] Therefore, how to improve the accuracy of UAV path planning has become an urgent problem to be solved. Summary of the invention

[0005] The embodiments of the present application provide a drone path planning method, a planning device, an electronic device, and a storage medium, which aim to solve the problem of low accuracy in drone path planning.

[0006] In a first aspect, an embodiment of the present application provides a method for path planning of a UAV, the method comprising: obtaining path planning information, the path planning information comprising the starting point, the end point and obstacle information of the UAV; based on the path planning information, using a rapidly expanding random tree algorithm to determine a feasible path; converting the feasible path into multiple candidate paths represented in the form of chromosomes in a genetic algorithm; based on each of the candidate paths, using the genetic algorithm to determine the optimal path.

[0007] In a possible implementation, the method of determining a feasible path based on the path planning information and using a fast expanding random tree algorithm includes: taking the starting point of the drone as the first nearest neighbor point of a search tree set, and determining a target sampling point according to a sampling point selection model; determining a second nearest neighbor point based on the Euclidean distance between the target sampling point and each node in the search tree set; determining a newly added node based on the path planning information, the target sampling point, the second nearest neighbor point and a preset expansion step; and determining the feasible path based on the newly added node when a first preset condition is met.

[0008] In a possible implementation, the method of determining a newly added node based on the path planning information, the target sampling point, the second nearest neighbor point and a preset extension step length by using an artificial potential field algorithm includes: respectively calculating the gravitational forces of the target sampling point and the endpoint of the drone on the second nearest neighbor point to obtain a first gravitational force and a second gravitational force; calculating the repulsive force of the target obstacle on the second nearest neighbor point based on the obstacle information and a repulsive force calculation formula, wherein the target obstacle is an obstacle closest to the second nearest neighbor point; calculating a resultant force based on the first gravitational force, the second gravitational force and the repulsive force; determining the forward component of the drone in different directions based on the second nearest neighbor point, the resultant force and the preset extension step length; and determining the newly added node based on the second nearest neighbor point and the forward component of the drone in different directions.

[0009] In a possible implementation, the repulsive force calculation formula includes:

[0010]

[0011] in, represents the direction vector between the second nearest neighbor point and the target obstacle, represents the direction vector between the second nearest neighbor point and the end point of the drone, p,p goal ,p obs represent the second nearest neighbor point, the endpoint of the UAV, and the position of the target obstacle respectively; n is a positive integer; ρ(p, p obs ) represents the Euclidean distance between the second nearest neighbor point and the target obstacle; ρ(p, p goal ) represents the Euclidean distance between the second nearest neighbor point and the endpoint of the drone, k r is the repulsive field gain constant, ρ0 represents the maximum safe distance between the second nearest neighbor point and the threat range, U rep (p) is the repulsive field function.

[0012] In a possible implementation, the sampling point selection model is:

[0013]

[0014] Among them, q rand represents the target sampling point, rand() is a random number between 0 and 1, δ is the preset target bias threshold, q goal Indicates the endpoint of the drone.

[0015] In a possible implementation, the method of determining the best path based on each of the candidate paths using the genetic algorithm includes: evaluating each of the candidate paths using a fitness function to determine a first path set, wherein the fitness function includes a path length function, a path smoothing function, and a path safety function; performing crossover and mutation operations on the first path set, and using an elite retention strategy to determine a second path set; evaluating the second path set using the fitness function to determine a fitness score corresponding to each path in the second path set; and determining the best path based on the fitness score when a second preset condition is met.

[0016] In a possible implementation, after determining the best path based on each of the candidate paths by using the genetic algorithm, the method further includes: smoothing the best path by using a cubic spline method.

[0017] In the second aspect, an embodiment of the present application provides a UAV path planning device, which includes: an information acquisition module for acquiring path planning information, the path planning information including the starting point, end point and obstacle information of the UAV; a feasible path determination module for determining a feasible path based on the path planning information using a rapidly expanding random tree algorithm; a data conversion module for converting the feasible path into multiple candidate paths represented in the form of chromosomes in a genetic algorithm; and an optimal path determination module for determining an optimal path based on each of the candidate paths using the genetic algorithm.

[0018] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method described in the first aspect or any one of the implementation methods thereof is implemented.

[0019] In a fourth aspect, an embodiment of the present application 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 method described in the first aspect or any one of the implementation methods thereof is implemented.

[0020] In a fifth aspect, an embodiment of the present application provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the method described in the first aspect or any one of the implementation methods thereof.

[0021] Compared with the prior art, the embodiments of the present application have the following beneficial effects: first, based on the acquired UAV path planning information, a fast expanding random tree algorithm is used to determine a feasible path; then, the feasible path is converted into multiple candidate paths represented in the form of chromosomes in a genetic algorithm, and each candidate path is processed by a genetic algorithm to determine the best path. Compared with the use of a single algorithm for path planning, the fast exploration capability of the fast expanding random tree algorithm and the global optimization capability of the genetic algorithm are combined to achieve rapid generation and fine adjustment of the path, enhance the smoothness and drivability of the planned path, make the final planned path better, and improve the efficiency and accuracy of UAV path planning in complex environments.

[0022] It can be understood that the drone path planning device, electronic device, computer-readable storage medium and computer program product provided in the embodiments of the present application have the same beneficial effects as the above-mentioned drone path planning method, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] 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.

[0024] Figure 1 A schematic diagram of a flow chart of a drone path planning method provided in one embodiment of the present application;

[0025] Figure 2 It is a flowchart of a RRT algorithm;

[0026] Figure 3 A schematic diagram of force analysis of a second nearest neighbor point;

[0027] Figure 4 A schematic diagram of path smoothness provided in one embodiment of the present application;

[0028] Figure 5 A structural block diagram of a drone path planning device provided in one embodiment of the present application;

[0029] Figure 6 A schematic diagram of the structure of an electronic device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0030] 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.

[0031] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] Under the current technological environment, the autonomous obstacle avoidance capability of drones mainly relies on onboard sensors and specific path planning algorithms. After the drone is started, it collects environmental data around it through sensors and transmits the environmental data to the onboard computer system. The computer system is responsible for analyzing and processing environmental data and planning a safe flight route in combination with the path planning algorithm. Subsequently, the computer system converts the planned path into specific flight instructions, guiding the drone to fly autonomously along the planned path until it reaches its destination.

[0037] The accuracy of autonomous obstacle avoidance of drones is affected by many factors, including the processing performance of the computer system, the accuracy of the sensor, the accuracy of path planning, the stability of communication, and external environmental conditions. Among them, the performance of the computer system, the quality of the sensor, and the accuracy of the path planning algorithm are key factors that can be optimized and adjusted manually. The accuracy of the path planned by the path planning algorithm and whether it is optimal have a great impact on the actual flight of the drone.

[0038] The current autonomous obstacle avoidance technology for drones generally relies on a single path planning algorithm to perform path planning to achieve automatic obstacle avoidance. The actual effect of this obstacle avoidance method is often not ideal. These problems are not only restricted by hardware conditions, but are also easily affected by changes in the external environment and the adaptability of the path planning algorithm to the existing environment, thus posing a severe challenge to the safe and autonomous flight of drones. Obviously, there is significant room for improvement in the current autonomous obstacle avoidance technology for drones, and it is urgent to improve its path planning effect and flight safety in complex environments through technological innovation.

[0039] In current drone systems, the existing single path planning algorithms used have limitations due to algorithm characteristics, resulting in poor paths planned by drones in different environments, which is not conducive to drone obstacle avoidance and further reduces the accuracy of drone autonomous obstacle avoidance.

[0040] Traditional path planning algorithms often have problems such as large amount of calculation and slow convergence speed when facing complex environments or high-dimensional spaces. The RRT algorithm is widely used in the field of path planning due to its efficient exploration ability and good global search performance, but the path it generates may be tortuous and lacks smoothness. Genetic algorithm, as an optimization algorithm that simulates natural selection and genetic mechanism, is good at searching for global optimal solutions, but lacks local search efficiency. Therefore, combining the two, using RRT to quickly generate the initial path, and then optimizing and adjusting the path through the genetic algorithm, can effectively improve the efficiency and quality of path planning.

[0041] The RRT algorithm is an algorithm widely used in the field of path planning and robot navigation. It is known for its high efficiency and global search capabilities. It has the following advantages: The RRT algorithm explores the search space by random sampling, and theoretically can cover the entire state space to find a feasible path from the start point to the end point, even if the space is complex or high-dimensional; The RRT algorithm can usually find a solution in a shorter time, especially for high-dimensional or complex environments; The RRT algorithm has simple logic, is easy to understand and implement, and does not require prior knowledge of the topological structure of the search space; In infinite time, the RRT algorithm can find a feasible path from the start point to the end point with probability 1, as long as such a path exists; The RRT algorithm can be easily combined with other algorithms to improve the efficiency and quality of path planning.

[0042] However, it also has the following disadvantages: the path generated by the RRT algorithm may not be optimal and may even be very tortuous, because it mainly focuses on quickly finding a feasible solution rather than an optimal solution; the performance of the RRT algorithm depends largely on its parameter settings (such as step size, maximum depth of the tree, sampling strategy, etc.), and different parameters may lead to very different results; although the RRT algorithm is more efficient than the gridding method, it may still require a lot of computing resources and memory in high-dimensional spaces or applications requiring high precision; in some cases, the RRT algorithm may fall into local optimality, especially in narrow channels or dense obstacle environments, resulting in the inability to find the global optimal path; the path generated by the RRT algorithm is usually a broken line composed of a series of straight line segments, which may not be suitable for application scenarios that require smooth trajectories, such as robot arm motion planning scenarios.

[0043] Genetic Algorithm (GA) is an optimization algorithm inspired by natural selection and genetic mechanisms, and is widely used to solve various complex problems. Genetic algorithms have the following advantages: Genetic algorithms simulate the process of natural selection, conduct global searches in the solution space, and can explore multiple potential solutions, so that there is a greater probability of finding the global optimal solution or a solution close to the global optimal solution; Genetic algorithms do not rely on the specific domain knowledge of the problem and can be applied to various types of problems, including nonlinear, multi-peak, and non-continuous problems; Genetic algorithms are suitable for parallel computing because individuals in the population can be evaluated independently, which helps to improve computing efficiency, especially when running on large-scale parallel processors; Genetic algorithms can adapt to different problem requirements by adjusting parameters such as crossover rate, mutation rate, and selection strategy, and have strong flexibility; Even in the face of complex optimization problems, genetic algorithms can maintain a certain degree of stability and effectiveness because they use the diversity of the population to avoid falling into local optimality.

[0044] Genetic algorithms have the following disadvantages: genetic algorithms usually require a large number of iterations to converge to a better solution, especially when dealing with high-dimensional or complex problems, and consume a lot of computing resources; the performance of genetic algorithms is highly dependent on parameter settings (such as population size, crossover rate, mutation rate, etc.), and inappropriate parameters may cause the algorithm to converge prematurely or fall into a local optimum; compared with some gradient descent methods, genetic algorithms converge more slowly, especially when approaching the optimal solution, requiring more iterations for fine-tuning; although genetic algorithms perform well in practice, their theoretical analysis is relatively complex, and it is difficult to predict their performance and convergence for certain specific problems; because genetic algorithms are based on randomness, the results of each run may be different, and sometimes poor solutions may be obtained, especially when the population diversity is insufficient.

[0045] Based on the above introduction, in order to solve the problems existing in the prior art, the present application proposes a UAV path planning method, which obtains path planning information, the path planning information includes the starting point, end point and obstacle information of the UAV; based on the path planning information, a fast expanding random tree algorithm is used to determine a feasible path; the feasible path is converted into multiple candidate paths represented in the form of chromosomes in a genetic algorithm; based on each candidate path, a genetic algorithm is used to determine the optimal path, thereby improving the accuracy of the UAV path planning.

[0046] For ease of understanding, the technical solution of the present application will be described in detail below with reference to the accompanying drawings.

[0047] Figure 1 The following is a flow chart of a method for drone path planning provided in one embodiment of the present application. For ease of explanation, only the part related to the present embodiment is shown. The method provided in the present embodiment includes the following steps:

[0048] S110, obtaining path planning information, where the path planning information includes the starting point, end point, and obstacle information of the UAV.

[0049] In a specific implementation, the drone's radar, camera and other sensors are used to obtain path planning information such as the starting point, end point and obstacle information in the target area where the drone is traveling. The obstacle information includes the location information of each obstacle in the target area where the drone is traveling.

[0050] As an example, drones include unmanned vehicles, unmanned ships, etc.

[0051] S120, based on the path planning information, a fast expanding random tree algorithm is used to determine a feasible path.

[0052] In the specific implementation, the search space is first constructed through the path planning information, and the RRT parameters (such as step size, maximum depth of the search tree, etc.) are initialized; then starting from the starting point of the UAV, the RRT algorithm is used to randomly expand the search tree until the end point of the UAV is reached or the preset number of iterations is reached to generate a feasible path.

[0053] S130, converting the feasible path into multiple candidate paths represented in the form of chromosomes in a genetic algorithm.

[0054] Specifically, the feasible paths generated by the RRT algorithm are converted into a chromosome representation suitable for genetic algorithm processing, and each chromosome represents a candidate path.

[0055] In the specific implementation, we first define the problem and the encoding scheme, determine the search space of the problem based on the feasible paths generated by the RRT algorithm, that is, all possible paths, and select a suitable encoding method to represent these paths. Common encoding methods include binary encoding, real number encoding, etc.; then initialize the population, and randomly generate a set of initial solutions (that is, candidate paths) based on all the possible paths mentioned above as the initial population of the genetic algorithm, and each candidate path is represented as a chromosome.

[0056] S140, based on each candidate path, using a genetic algorithm to determine the best path.

[0057] Specifically, the genetic algorithm mainly includes the following operations:

[0058] Select fitness function: define a fitness function to evaluate the quality of each chromosome (i.e. candidate path) and obtain the fitness value. The fitness function can reflect the quality or performance of the path.

[0059] Selection operation: select some chromosomes for reproduction according to the fitness value. Common selection methods include roulette selection, tournament selection, etc.

[0060] Crossover operation: Perform a crossover operation on the selected chromosomes to generate new chromosomes (offspring). The purpose of the crossover operation is to combine the excellent characteristics of the parent generation to generate better offspring.

[0061] Mutation operation: Perform mutation operation on the newly generated chromosomes to increase the diversity of the population. Mutation operation can prevent the genetic algorithm from falling into the local optimal solution;

[0062] Iterative evolution: Repeat the selection, crossover, and mutation operations until the termination condition is met.

[0063] In a specific implementation, an initial population is generated based on each candidate path, the fitness of each candidate path in the population is calculated by a fitness function, a path is selected from the candidate paths for reproduction according to the calculated fitness, crossover and mutation operations are performed on the selected path to generate new offspring, and a new generation of population is generated according to a certain strategy (such as an elite strategy), and it is determined whether the preset termination conditions are met, such as whether the maximum number of iterations has been reached, whether the quality of the solution meets the requirements, or whether the fitness is no longer significantly improved, etc. When the preset termination conditions are met, the path with the highest fitness in the current population is output as the optimal path.

[0064] The technical solution provided in this embodiment first determines a feasible path based on the acquired UAV path planning information using a rapidly expanding random tree algorithm; then converts the feasible path into multiple candidate paths represented in the form of chromosomes in a genetic algorithm, and uses a genetic algorithm to process each candidate path to determine the best path. Compared with using a single algorithm for path planning, it combines the rapid exploration capability of the rapidly expanding random tree algorithm and the global optimization capability of the genetic algorithm, realizes the rapid generation and fine adjustment of the path, enhances the smoothness and drivability of the planned path, makes the final planned path better, and improves the efficiency and accuracy of UAV path planning in complex environments.

[0065] Based on the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, based on the path planning information, a fast expanding random tree algorithm is used to determine a feasible path, including:

[0066] The starting point of the drone is used as the first nearest neighbor point of the search tree set, and the target sampling point is determined according to the sampling point selection model;

[0067] Determine the second nearest neighbor based on the Euclidean distance between the target sampling point and each node in the search tree set;

[0068] Based on the path planning information, target sampling points, second nearest neighbor points and preset expansion step length, the artificial potential field algorithm is used to determine the newly added nodes;

[0069] When the first preset condition is met, a feasible path is determined based on the newly added node.

[0070] Specifically, the RRT algorithm constructs a search tree by random sampling in space. The search tree expands from the starting point to the end point of the UAV, and the performance constraints of the UAV should be met during the node expansion process.

[0071] In the specific implementation, the initialization setting is first performed, the starting point of the drone is used as the first nearest neighbor point of the search tree set, and the maximum number of iterations is determined. Then, the sampling point selection model is used to determine the target sampling point. According to the Euclidean distance between the target sampling point and each node in the search tree set, the node with the closest Euclidean distance to the target sampling point is selected as the second nearest neighbor point. Based on the path planning information, the target sampling point, the second nearest neighbor point and the preset expansion step size, the artificial potential field algorithm is used to determine the newly added node. If the connection line between the second nearest neighbor point and the newly added node passes through an obstacle or the newly added node is in the obstacle, it is necessary to reselect the newly added node, otherwise the newly added node is stored in the search tree set. When the first preset condition is met, the algorithm terminates, and a feasible path is determined based on the newly added node.

[0072] As an example, the first preset condition includes when the newly added node reaches the end point of the drone or the current number of iterations reaches the maximum number of iterations.

[0073] As an example, the sampling point selection model is:

[0074]

[0075] Among them, q rand represents the target sampling point, rand() is a random number between 0 and 1, δ is the preset target bias threshold, q goal Indicates the end point of the drone. When the rand() value is greater than the preset target bias threshold, a random point is selected in the sampling space as the target sampling point, otherwise the end point of the drone is used as the target sampling point.

[0076] As an example, the process of the RRT algorithm is as follows Figure 2 As shown, first perform initialization settings to determine whether the current number of iterations is less than or equal to the maximum number of iterations M. If not, it means that the maximum number of iterations has been reached and the process ends; if so, random sampling points q are generated in sequence rand , determine the nearest neighbor point q near , generate a new node q new , and add 1 to the current iteration number to determine the new node q new Whether it collides with an obstacle. If not, return to the step of determining whether the current number of iterations is less than or equal to the maximum number of iterations M to continue the process; if so, add the new node q new Join the search tree set to further determine the new node q new Is it the end point of the drone? If so, end the process; if not, return to the step of determining whether the current number of iterations is less than or equal to the maximum number of iterations M to continue the process.

[0077] The technical solution provided in this embodiment guides the generation of target sampling points by target bias, thereby improving the search efficiency of the RRT algorithm, improving the accuracy of the target sampling points, and further improving the accuracy of the feasible path.

[0078] On the basis of the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, based on the path planning information, the target sampling point, the second nearest neighbor point and the preset expansion step, an artificial potential field algorithm is used to determine the newly added node, including:

[0079] Calculate the gravitational forces of the target sampling point and the end point of the UAV on the second nearest neighbor point respectively, and obtain the first gravitational force and the second gravitational force;

[0080] Based on the obstacle information and the repulsion calculation formula, the repulsion of the target obstacle on the second nearest neighbor point is calculated, and the target obstacle is the obstacle closest to the second nearest neighbor point;

[0081] Calculate the resultant force based on the first attractive force, the second attractive force and the repulsive force;

[0082] Determine the forward component of the UAV in different directions based on the second nearest neighbor point, the resultant force and the preset expansion step size;

[0083] Based on the second nearest neighbor point and the forward component of the UAV in different directions, the newly added node is determined.

[0084] Specifically, this embodiment optimizes the generation method of new nodes by adding an artificial potential field method on the basis of the original RRT algorithm to accelerate the convergence speed of the algorithm. The second nearest neighbor point is taken as point P, then the original gravitational potential field function U att (p), the original repulsive potential field function U rep (p) and the original resultant potential field function U total Respectively expressed as

[0085]

[0086] U total =∑U rep +∑U att

[0087] Among them, k a is the gravitational field gain constant: k r is the repulsive field gain constant; ρ0 is the maximum safe distance between the second nearest neighbor point and the threat range, which can be preset according to the actual situation; ρ g (p) and ρ(p) represent the Euclidean distance between the second nearest neighbor and the end point of the drone and the Euclidean distance between the second nearest neighbor and the nearest threat center, respectively. The magnitudes of gravity and repulsion are the negative gradients of the gravitational field function and the repulsive field function, respectively. From the above formula, we can get the original gravity calculation formula Fatt (p), original repulsive force calculation formula F rep (p) and the original resultant force calculation formula F total They are:

[0088] F att (p) = k a ρ g (p)

[0089]

[0090] F total =∑F rep +∑F att

[0091] By combining the artificial potential field method, the end point and target sampling point of the drone generate gravitational force on the second nearest neighbor point, and the target obstacle generates repulsive force on the second nearest neighbor point. The combined force of the three determines the generation direction of the new node.

[0092] As an example, a force condition of the second nearest neighbor point is as follows Figure 3 shown. Figure 3 Medium, F att1 and F att2 The target sampling points q are rand and the drone's endpoint q goal For the second nearest neighbor point q near The gravitational force, F rep is the repulsive force of the obstacle on the second nearest neighbor. According to the parallelogram rule, the red solid line is F att1 With F rep The direction of the resultant force, the blue solid line is composed of the red solid line and F att2 The vector sum is the direction of the resultant force on the second nearest neighbor, that is, the expansion direction of the newly added node. However, when the drone is close to the obstacle, the repulsive force it is subjected to will become very large. When the repulsive force is greater than the gravitational force, the drone will not be able to reach the end point. Therefore, the potential field function is improved, and the improved potential field function is:

[0093]

[0094] Among them, p,p goal ,p obs They represent the second nearest neighbor, the end point of the drone, and the position of the target obstacle respectively; n is a positive integer; ρ(p, p obs ) represents the Euclidean distance between the second nearest neighbor and the target obstacle; ρ(p, p goal ) represents the Euclidean distance between the second nearest neighbor and the destination of the drone, k ris the repulsive field gain constant, ρ0 represents the maximum safe distance between the second nearest neighbor point and the threat range, which can be preset according to the actual situation; U rep (p) is the repulsive field function, U att (p) is the gravitational field function; when the UAV approaches the target obstacle, the repulsive force generated by the target obstacle will increase with ρ(p, p goal ) decreases, thus avoiding the phenomenon that the repulsive force is greater than the attractive force. Based on the improved potential field function, the repulsive force calculation formula is obtained as follows:

[0095]

[0096] in, represents the direction vector between the second nearest neighbor point and the target obstacle, Represents the direction vector between the second nearest neighbor and the drone’s destination.

[0097] In the specific implementation, the gravitational forces of the target sampling point and the end point of the drone on the second nearest neighbor are calculated respectively, and then the repulsive force of the target obstacle on the second nearest neighbor is calculated according to the above repulsive force calculation formula. Then, the resultant force F calculated by each gravitational force and repulsive force is total Decompose into the force components in the x-axis and y-axis directions, respectively expressed as F x ,F y Indicates; Assuming the coordinates of the second nearest neighbor are (xc, yc), the forward component of the drone in different directions is determined by the component of the resultant force on the second nearest neighbor on the x and y axes. The absolute value of the maximum value of the resultant force component on each coordinate axis is selected as F max , then the forward component of the UAV in different directions can be determined as

[0098]

[0099] Among them, l is the preset expansion step size, l x and l y are the forward components of the UAV on the x and y axes respectively; k is the proportional coefficient, which is calculated from the maximum value of the resultant force component F max In this way, the coordinates of the newly added node can be obtained. Let the coordinates of the newly added node be (x new ,y new ), then the coordinate equation of the newly added node is

[0100]

[0101] The technical solution provided in this embodiment, combined with the artificial potential field method, improves the generation method of new nodes, greatly reduces the number of redundant points, improves the search efficiency and accuracy of the RRT algorithm, improves the accuracy of new nodes, and further improves the accuracy of feasible paths.

[0102] Based on the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, based on each candidate path, a genetic algorithm is used to determine the best path, including:

[0103] Using a fitness function to evaluate each candidate path to determine a first path set, the fitness function includes a path length function, a path smoothing function and a path safety function;

[0104] Performing crossover and mutation operations on the first path set and adopting an elite retention strategy to determine the second path set;

[0105] The second path set is evaluated using a fitness function to determine a fitness score corresponding to each path in the second path set;

[0106] When the second preset condition is met, the best path is determined based on the fitness score.

[0107] In the specific implementation, the fitness function is first used to evaluate the quality of each candidate path, and the excellent candidate paths are selected according to the evaluation results to form the first path set and enter the next generation; some paths are selected from the first path set for crossover operations to generate new paths to explore a better solution space; slight perturbations are implemented on the selected paths to increase population diversity and avoid local optimality; an elite retention strategy is used to directly retain a certain proportion of excellent paths to form the second path set to the next generation to ensure the inheritance of excellent genes; then the fitness function is used again to evaluate the fitness of the paths in the second path set, and the fitness of the new generation of paths is recalculated to obtain the fitness scores corresponding to each path in the second path set; when the second preset condition is met, the path with the highest fitness score in the second path set is determined as the best path.

[0108] As an example, the fitness function f(p) can be expressed as:

[0109]

[0110] Among them, α, β and γ represent the weight factors of path length, path smoothness and path safety respectively, L(p) represents the path length function, S(p) represents the path smoothness function, and R(p) represents the path safety function.

[0111] Exemplarily, the path length function can be expressed as:

[0112]

[0113] Among them, (x i ,y i ) represents the coordinates of the ith track point, and m is the total number of track points.

[0114] For example, the path smoothness is Figure 4 As shown, q start and q goal Respectively represent the starting point and end point of the drone, and the path smoothing function can be expressed as:

[0115]

[0116] Among them, the track segment and The angle between i-1 ,l i ).

[0117] Exemplarily, the path safety function can be expressed as:

[0118]

[0119] Among them, D i Represents the distance between the ith track point and the nearest obstacle.

[0120] As an example, the second preset condition includes reaching a preset number of iterations or the fitness improvement is not obvious.

[0121] As an example, when performing the crossover operation and the mutation operation on the first path set, the crossover probability is set to 0.06 and the mutation probability is set to 0.01.

[0122] The technical solution provided in this embodiment uses a genetic algorithm to further optimize the feasible path generated by the RRT algorithm, thereby improving the accuracy of the optimal path.

[0123] Based on the above embodiment, this embodiment further illustrates and optimizes the technical solution. Specifically, in this embodiment, after the optimal path is determined based on each candidate path by using a genetic algorithm, the following is further included:

[0124] The optimal path is smoothed using cubic splines.

[0125] The technical solution provided in this embodiment uses the cubic spline method to smooth the optimal path, thereby enhancing the smoothness and drivability of the optimal path and further improving the accuracy of the optimal path.

[0126] It should be noted that the technical solution provided in this application has been proven to be feasible and effective through theoretical analysis, simulation, and physical construction.

[0127] Figure 5 This is a structural block diagram of a drone path planning device provided by an embodiment of the present application. For the sake of convenience, only the parts related to the embodiment of the present application are shown. Figure 5 The UAV path planning device 500 may include an information acquisition module 501, a feasible path determination module 502, a data conversion module 503 and an optimal path determination module 504.

[0128] The information acquisition module 501 is used to acquire path planning information, which includes the starting point, end point and obstacle information of the drone.

[0129] The feasible path determination module 502 is used to determine a feasible path based on the path planning information by using a fast expanding random tree algorithm.

[0130] The data conversion module 503 is used to convert the feasible path into multiple candidate paths represented in the form of chromosomes in the genetic algorithm.

[0131] The best path determination module 504 is used to determine the best path based on each candidate path by using a genetic algorithm.

[0132] An unmanned aerial vehicle path planning device provided in an embodiment of the present application has the same beneficial effects as the above-mentioned unmanned aerial vehicle path planning method.

[0133] 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.

[0134] 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.

[0135] Figure 6 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Figure 6 As shown, the electronic device 6 of this embodiment includes: at least one processor 60 ( Figure 6Only one is shown in the figure), a memory 61, and a computer program 62 stored in the memory 61 and executable on at least one processor 60, the processor 60 executes the computer program 62 to implement the above Figure 1 The steps in the method embodiment, or the implementation of the above Figure 5 Functions of each module / unit in the device embodiment.

[0136] The electronic device 6 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The electronic device 6 may include but is not limited to a processor 60 and a memory 61. Those skilled in the art will appreciate that Figure 6 It is only an example of the electronic device 6 and does not constitute a limitation on the electronic device 6. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0137] The processor 60 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.

[0138] In some embodiments, the memory 61 may be an internal storage unit of the electronic device 6, such as a hard disk or memory of the electronic device 6. In other embodiments, the memory 61 may also be an external storage device of the electronic device 6, such as a plug-in hard disk, a smart memory card (SmartMedia Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 6. Further, the memory 61 may also include both an internal storage unit of the electronic device 6 and an external storage device. The memory 61 is used to store an operating system, an application program, a boot loader (BootLoader), data, and other programs, such as program codes of a computer program, etc. The memory 61 may also be used to temporarily store data that has been output or is to be output.

[0139] The embodiment of the present application further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0140] 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, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. 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 include at least: any entity or device capable of carrying the computer program code to an electronic device, a recording medium, a computer memory, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), an electric carrier signal, a telecommunication signal, and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0141] A computer-readable storage medium provided in an embodiment of the present application has the same beneficial effects as the above-mentioned drone path planning method.

[0142] An embodiment of the present application provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0143] A computer program product provided in an embodiment of the present application has the same beneficial effects as the above-mentioned drone path planning method.

[0144] 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.

[0145] 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.

[0146] In the embodiments provided in the present application, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0147] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0148] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for UAV path planning, characterized in that: The method comprises: Obtaining path planning information, wherein the path planning information includes the starting point, end point, and obstacle information of the drone; Based on the path planning information, a fast expanding random tree algorithm is used to determine a feasible path; Converting the feasible path into multiple candidate paths represented in the form of chromosomes in a genetic algorithm; Based on each of the candidate paths, the optimal path is determined using the genetic algorithm.

2. The method according to claim 1, characterized in that The method of determining a feasible path based on the path planning information by using a fast expanding random tree algorithm includes: The starting point of the drone is used as the first nearest neighbor point of the search tree set, and the target sampling point is determined according to the sampling point selection model; Determine the second nearest neighbor point based on the Euclidean distance between the target sampling point and each node in the search tree set; Based on the path planning information, the target sampling point, the second nearest neighbor point and the preset expansion step, an artificial potential field algorithm is used to determine a new node; When the first preset condition is met, the feasible path is determined based on the newly added node.

3. The method according to claim 2, characterized in that The method of determining a newly added node by using an artificial potential field algorithm based on the path planning information, the target sampling point, the second nearest neighbor point and a preset expansion step size includes: Calculate the gravitational forces of the target sampling point and the endpoint of the drone on the second nearest neighbor point respectively to obtain a first gravitational force and a second gravitational force; Based on the obstacle information and the repulsion calculation formula, calculate the repulsion of the target obstacle on the second nearest neighbor point, the target obstacle being the obstacle closest to the second nearest neighbor point; Calculating a resultant force based on the first gravitational force, the second gravitational force, and the repulsive force; Determining the forward component of the UAV in different directions based on the second nearest neighbor point, the resultant force and the preset extension step size; The newly added node is determined based on the second nearest neighbor point and the forward component of the drone in different directions.

4. The method according to claim 3, characterized in that The repulsive force calculation formula includes: in, represents the direction vector between the second nearest neighbor point and the target obstacle, represents the direction vector between the second nearest neighbor point and the end point of the drone, p,p goal ,p obs represent the second nearest neighbor point, the endpoint of the UAV, and the position of the target obstacle respectively; n is a positive integer; ρ(p, p obs ) represents the Euclidean distance between the second nearest neighbor point and the target obstacle; ρ(p, p goal ) represents the Euclidean distance between the second nearest neighbor point and the endpoint of the drone, k r is the repulsive field gain constant, ρ0 represents the maximum safe distance between the second nearest neighbor point and the threat range, U rep (p) is the repulsive field function.

5. The method according to claim 2, characterized in that: The sampling point selection model is: Among them, q rand represents the target sampling point, rand() is a random number between 0 and 1, δ is the preset target bias threshold, q goal Indicates the endpoint of the drone.

6. The method according to claim 1, characterized in that The step of using the genetic algorithm to determine the best path based on each of the candidate paths includes: Using a fitness function to evaluate each of the candidate paths to determine a first path set, the fitness function comprising a path length function, a path smoothing function, and a path safety function; Performing a crossover operation and a mutation operation on the first path set, and adopting an elite retention strategy to determine a second path set; Using the fitness function to evaluate the second path set, and determining a fitness score corresponding to each path in the second path set; When the second preset condition is met, the optimal path is determined based on the fitness score.

7. The method according to any one of claims 1 to 6, characterized in that: After determining the best path based on each of the candidate paths by using the genetic algorithm, the method further includes: The optimal path is smoothed using a cubic spline method.

8. A drone path planning device, characterized in that: The device comprises: An information acquisition module is used to acquire path planning information, wherein the path planning information includes the starting point, end point and obstacle information of the UAV; A feasible path determination module, used to determine a feasible path based on the path planning information using a fast expanding random tree algorithm; A data conversion module, used for converting the feasible path into a plurality of candidate paths represented in the form of chromosomes in a genetic algorithm; The optimal path determination module is used to determine the optimal path based on each of the candidate paths using the genetic algorithm.

9. An electronic 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 steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images