Global path planning method and device for unmanned vehicle, and storage medium

Abstract

The invention discloses a global path planning method and device for an unmanned vehicle, and a storage medium, and the method comprises the following steps: obtaining topographic information and soil information, and representing the topographic condition of a target region; setting a maneuverability model and a path planning algorithm RRT * algorithm; performing multi-mode reliability analysis on the unmanned vehicle under the uncertain terrain condition; the reliability constraints are integrated into a path planning algorithm based on RRT *, and the shortest path meeting the two reliability constraints is recognized. The method is used for coping with two maneuverability failure modes, namely the maximum reachable speed and the vehicle vertical acceleration, caused by terrain uncertainty. The algorithm can ensure that the vehicle can find the optimal path in the complex terrain and can safely and reliably complete the task. The method has the advantages of simplicity, high efficiency, high reliability and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned vehicle path planning, and in particular to a global path planning method, device and storage medium for an unmanned vehicle. Background Art

[0002] With the development of autonomous ground vehicle technology, especially in the field of unmanned vehicles and robotics, path planning has become a key issue to ensure the efficient and safe operation of these vehicles in complex terrains. Traditional path planning methods mainly focus on minimizing travel time or path length. However, in uncertain and rugged terrain conditions, unmanned ground vehicles may face operational failures due to terrain changes, such as getting stuck in mud or being unable to pass steep slopes. Therefore, a planning method is needed that can not only generate a suitable path, but also ensure the overall reliability and safety of the path in complex terrain.

[0003] The Chinese patent publication number CN118500429A "Path Planning Method and System Based on Improved Artificial Potential Field RRT*" mainly provides a path planning method and system based on improved artificial potential field RRT*. The method randomly samples sampling points in the sampling space based on the RRT* algorithm, corrects the sampling points with the improved artificial potential field, and obtains the new node position with an adaptive step size, overcoming the problems of strong randomness and excessive invalid expansion of the native RRT* algorithm. The method has a simple structure, lightweight operation, and improves the rate of planning paths, which can improve the adaptability to different environments and accelerate the convergence speed.

[0004] The above path planning methods ignore the uncertainty of terrain conditions when dealing with complex terrain, such as changes in soil properties, uneven slopes, etc., which may cause the autonomous driving vehicle to fail in actual operations, such as getting stuck in the mud or losing stability. Summary of the invention

[0005] The present invention provides a global path planning method, device and storage medium for an unmanned vehicle, which can solve at least one of the technical problems in the background technology.

[0006] To achieve the above object, the present invention adopts the following technical solutions: A global path planning method for an unmanned vehicle comprises the following steps: Obtain terrain information and soil information to characterize the terrain conditions of the target area; Set the mobility model and path planning algorithm RRT* algorithm; Multimodal reliability analysis of unmanned vehicles under uncertain terrain conditions; The reliability constraints are integrated into the RRT*-based path planning algorithm to identify the shortest path that satisfies the two reliability constraints.

[0007] Furthermore, the acquisition of terrain information and soil information to characterize the terrain conditions of the target area includes the following steps: The target mapping is called the configuration space and is represented by Indicates; In the configuration space, the space with obstacles is called obstacle space, denoted by ; The space in which the unmanned vehicle can complete its mission is called free space. express; Any position in configuration space has two dimensions, given by given; The starting point is also called the initial point. Indicates that the end point is the goal to be achieved, with Represents; the connection between any nodes is called an edge; the planning path of the task is defined by Γ.

[0008] Further, setting the mobility model and the path planning algorithm RRT* algorithm includes the following steps, The maneuverability of a vehicle is closely related to vehicle dynamics and terrain conditions; the generalized formula for the maneuverability model is as follows:

[0009] in It means the vehicle is in position The output of mobility, is a general mobility model, is the vector of vehicle related parameters, is a vector representing the conditions around the vehicle; The path planning algorithm RRT* is obtained by The iteration of the growing tree structure begins to find the optimal path; a random node is generated using a sampling strategy ;if , is rejected; the sampling process continues until there is a random sample in the free space, i.e. ; Then identify the tree T The nearest node ;if Can be Access, which means and There is no obstacle between By connecting and is inserted into the tree; otherwise, a new node is generated using the redirect function and by connecting it to Add to tree; perform collision checking to ensure and There are no obstacles between them; After passing the collision check, the tree is modified using the rewiring operation; the above operations are performed iteratively until the optimal path is found; In the mobility model, the mobility model is combined with the path planning algorithm RRT* to identify the optimal path while satisfying the physical constraints.

[0010] Furthermore, the multimodal reliability analysis of unmanned vehicles under uncertain terrain conditions includes: S3.1. Sources of uncertainty and modeling The uncertainty comes from terrain and soil related parameters, and Gaussian random fields are used to model the uncertainty of the parameters; the spatial correlation between any two locations on the map is described by the autocorrelation function as follows:

[0011] In the formula is the relevant length along the space; , are any two points in the graph, is the correlation between these two points; In order to generate random realizations of the random field for reliability analysis, the truncated Karhunen-Loeve expansion is adopted as follows:

[0012] in is a sample of generated realizations of a random field; is the spatial coordinate of the random field represents the mean function; M is the cutoff number; and are the eigenvalues ​​and eigenvectors obtained by eigenanalysis of the correlation matrix; and are independent standard Gaussian random variables; S3.2 Reliability analysis of maximum achievable speed S3.2.1 The reliability of the maximum achievable speed is defined as follows:

[0013] in is the maximum achievable speed, is the speed threshold, ” stands for “for all”, and is the spatial location The terrain and soil related parameter vectors are: ; S3.2.2. Given the training data, the initial Gaussian process model GP is obtained in the high-fidelity simulation model as:

[0014] in is the maximum achievable speed predicted by GP; and are the vectors representing the basis functions and the corresponding coefficients respectively; is the error term; In addition to the training data, a set of input samples is generated using the Monte Carlo simulation method MCS, which is iteratively improved on the GP model given above; in each iteration, the identified new training samples are added to the training set to improve the GP model; after the algorithm converges, the safe area and failure area of ​​the vehicle's maximum achievable speed are determined; MCS can be used to calculate the vehicle's mission maneuverability reliability expressed in terms of the maximum achievable speed:

[0015] S3.3 Reliability analysis of vertical acceleration S 3.3.1. The vertical acceleration is insensitive to soil parameters, and the effect of soil type on vertical acceleration is ignored; the vertical acceleration model is simplified as a function of terrain elevation as follows:

[0016] in For spatial location The vertical acceleration at ( )for The elevation of and is a nonlinear function, and the model is also called the nonlinear autoregressive exogenous model NARX; S3.3.2. First, the Gaussian mixture model GMM is used to segment multiple NARX models of different segments from the entire prediction area. Then, the NARX models are dynamically integrated on the path to predict the vertical acceleration along the path. The reliability of the vertical acceleration is as follows:

[0017] in,

[0018] is the acceleration predicted by the NARX model along the planned path; Vertical acceleration threshold.

[0019] Furthermore, the RRT algorithm is combined with the agent model to explore the path and find the optimal path, including: S4.1. Considering these two failure modes in unmanned vehicle path planning, the reliability-based path planning model is as follows:

[0020] st

[0021]

[0022]

[0023] In the formula To design the path, So To minimize the cost, and Mission maneuver reliability is expressed in terms of maximum achievable speed and vertical acceleration, respectively; is the reliability target in terms of maximum achievable speed, is the reliability target in terms of vertical acceleration; First, a tree represented by T is initialized, which contains all the nodes generated during the path exploration process; in the first step, there is only the starting point and goals In t; Next, randomly generate a node , and find the distance in the tree The nearest node ( ); it is noticed that the first random node, the closest node is ; The function used to find the nearest node is as follows:

[0024] After finding a random node, use a steering function to generate a new node , expressed as:

[0025] in represents the steering function; then arrive Reliability and Check; if the path meets the reliability target, find a neighbor node within the radius and record it as:

[0026] The radius is defined as:

[0027] in is a constant defined according to the environment, is the number of iterations, dim is the dimension of the configuration space; After that, at the lowest cost from Select The algorithm performs rewiring and rebuilds to The above operations are iterated until The area where the target point is reached.

[0028] On the other hand, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.

[0029] On the other hand, the present invention further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0030] As can be seen from the above technical solutions, the present invention aims at the shortcomings of traditional path planning methods in dealing with uncertain and rugged terrain, which easily causes the vehicle to get stuck or lose mobility. A path planning algorithm that comprehensively considers terrain uncertainty and multiple reliability constraints is proposed to deal with two mobility failure modes (maximum achievable speed and vehicle vertical acceleration) caused by terrain uncertainty. The algorithm can ensure that the vehicle can find the optimal path in complex terrain and complete the task safely and reliably. The present invention has the advantages of simplicity, high efficiency, and high reliability.

[0031] Compared with the prior art, the present invention has the following advantages: (1) The original RRT* algorithm ignores the uncertainty of terrain conditions and considers two reliability constraints caused by the uncertainty in terrain deformation.

[0032] (2) To address the problem that the high-fidelity simulation model is computationally too large, two alternative models were created for reliability analysis using limited high-fidelity simulation data to improve efficiency.

[0033] (3) Combining RRT* with the two reliability constraints of the agent model, a new planning method is obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a flowchart of the present invention. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0036] like Figure 1 As shown, the global path planning method for the unmanned vehicle described in this embodiment includes the following steps: Step 1: Obtain terrain information (height map) and soil information (soil map) to characterize the terrain conditions of the target area; Step 2: Mobility model and RRT* algorithm; Step 3: Multimodal reliability analysis of unmanned vehicles under uncertain terrain conditions; Step 4: Integrate the reliability constraints into the RRT*-based path planning algorithm to identify the shortest path that satisfies the two reliability constraints.

[0037] The following are specific instructions: Step 1: Obtain terrain information (height map) and soil information (soil map) to characterize the terrain conditions of the target area.

[0038] The target mapping is called the configuration space and is represented by In the configuration space, the space with obstacles is called obstacle space, which is denoted by The space in which the unmanned vehicle can complete its mission is called free space. Any position in the configuration space has two dimensions, represented by The starting point is also called the initial point, The end point is the goal to be achieved, indicated by The connection between any nodes is called an edge. The planning path of the task is defined by Γ.

[0039] Step 2: Mobility Model and RRT* Algorithm 2.1 The mobility of a vehicle is closely related to vehicle dynamics and terrain conditions. The generalized formula of the mobility model is as follows:

[0040] in It means the vehicle is in position The maneuverability outputs, e.g., maximum achievable speed and maximum vertical acceleration, is a general mobility model, is the vector of vehicle related parameters, is a vector representing the conditions surrounding the vehicle (e.g., slope and soil parameters).

[0041] 2.2 RRT* RRT* is obtained from The tree structure is iterated to find the optimal path. A random node is generated using a sampling strategy. .if , is rejected. The sampling process continues until there is a random sample in the free space; that is, Then identify the tree T The nearest node .if Can be Access, which means and There is no obstacle between By connecting and is inserted into the tree. Otherwise, a new node is generated using the redirect function and by connecting it to Add to the tree. Perform collision checking to ensure and There are no obstacles in between. After passing the collision check, the tree is modified using the rewiring operation. The above operations are performed iteratively until the optimal path is found.

[0042] 2.3 In the mobility model, the mobility model is combined with the path planning algorithm (RRT*) to identify the optimal path while satisfying the physical constraints.

[0043] Step 3: Multimodal reliability analysis of unmanned vehicles under uncertain terrain conditions 3.1 Uncertainty Sources and Modeling The uncertainty comes from terrain and soil related parameters, and Gaussian random fields are used to model the uncertainty of the parameters. The spatial correlation between any two locations on the map is described by the autocorrelation function as follows:

[0044] In the formula is the relevant length along the space; , are any two points in the graph, is the correlation between these two points.

[0045] In order to generate random realizations of the random field for reliability analysis, the truncated Karhunen-Loeve (KL) expansion is adopted as follows:

[0046] in is a generated realization (sample) of a random field; is the spatial coordinate of the random field represents the mean function; M is the cutoff number; and are the eigenvalues ​​and eigenvectors obtained by eigenanalysis of the correlation matrix; and are independent standard Gaussian random variables.

[0047] 3.2 Reliability analysis of maximum achievable speed 3.2.1 To operate successfully in an off-road environment, the vehicle needs to maintain a certain minimum speed to avoid getting stuck in deformable terrain. Based on this consideration, the reliability of the maximum achievable speed is defined as follows:

[0048] in is the maximum achievable speed, is the speed threshold, ” stands for “for all”, and is the spatial location The terrain and soil related parameters considered are

[0049] 3.2.2 Given the training data, the initial Gaussian process model (GP) obtained in the high-fidelity simulation model is:

[0050] in is the maximum achievable speed predicted by GP; and are the vectors representing the basis functions and the corresponding coefficients respectively; is the error term.

[0051] In addition to the training data, a set of input samples is generated using the Monte Carlo simulation method MCS, which is iteratively improved in the GP model given above. In each iteration, the identified new training samples are added to the training set to improve the GP model. After the algorithm converges, the safe area and failure area of ​​the maximum achievable speed of the vehicle are determined. MCS can be used to express the reliability of vehicle mission maneuvers in terms of the maximum achievable speed:

[0052] 3.3 Reliability analysis of vertical acceleration 3.3.1 The vertical acceleration is insensitive to soil parameters, and the effect of soil type on vertical acceleration can be ignored. The vertical acceleration model can be simplified as a function of terrain elevation as follows:

[0053] in For spatial location The vertical acceleration at ( )for The elevation of and is a nonlinear function. This model is also called nonlinear autoregressive exogenous model (NARX). 3.3.2 First, a Gaussian mixture model (GMM) is used to segment multiple NARX models of different segments from the entire prediction area. Then, the NARX models are dynamically integrated on the path to predict the vertical acceleration along the path. The reliability of the vertical acceleration is as follows:

[0055] in,

[0056] is the acceleration predicted by the NARX model along the planned path; Vertical acceleration threshold.

[0057] Step 4: Use the RRT algorithm combined with the agent model to explore the path and find the optimal path.

[0058] 4.1 Considering these two failure modes in unmanned vehicle path planning, the reliability-based path planning model is as follows:

[0059] st (constraint)

[0060]

[0061]

[0062] In the formula To design the path, So To minimize the cost, and Mission maneuver reliability is expressed in terms of maximum achievable speed and vertical acceleration, respectively. is the reliability target in terms of maximum achievable speed, is the reliability target in terms of vertical acceleration.

[0063] 4.2 Algorithm Description First, a tree T is initialized, which contains all the nodes generated during the path exploration process. In the first step, there is only the starting point and goals In t. Next, randomly generate a node , and find the distance in the tree The nearest node ( ). It is noticed that the first random node, the closest node is The function used to find the nearest node is as follows:

[0064] After finding a random node, use a steering function to generate a new node , expressed as:

[0065] in represents the steering function. Then arrive Reliability and If the path meets the reliability target, find a neighbor node within the radius and record it as:

[0066] The radius is defined as:

[0067] in are constants defined according to the environment, is the number of iterations, and dim is the dimension of the configuration space.

[0068] After that, at the lowest cost from Select The algorithm performs rewiring to rebuild The above operations are performed iteratively until The area where the target point is reached.

[0069] Compared with the prior art, the present invention has the following advantages: (1) The original RRT* algorithm ignores the uncertainty of terrain conditions and considers two reliability constraints caused by the uncertainty in terrain deformation.

[0070] (2) To address the problem that the high-fidelity simulation model is computationally too large, two alternative models were created for reliability analysis using limited high-fidelity simulation data to improve efficiency.

[0071] (3) Combining RRT* with the two reliability constraints of the agent model, a new planning method is obtained.

[0072] On the other hand, the present invention further discloses a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the above method.

[0073] On the other hand, the present invention further discloses a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0074] In another embodiment provided in the present application, a computer program product including instructions is also provided, which, when executed on a computer, enables the computer to execute the global path planning method for any unmanned vehicle in the above-mentioned embodiments.

[0075] It is understandable that the system, device and storage medium provided in the embodiments of the present invention correspond to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant contents can refer to the corresponding parts in the above methods.

[0076] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive Solid State Disk (SSD)), etc.

[0077] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0078] Each embodiment in this specification is described in a related manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0079] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, 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 invention.

Claims

1. A global path planning method for an unmanned vehicle, characterized in that: The following steps are included: Obtain terrain information and soil information to characterize the terrain conditions of the target area; Set the mobility model and path planning algorithm RRT* algorithm; Multimodal reliability analysis of unmanned vehicles under uncertain terrain conditions; The reliability constraints are integrated into the RRT*-based path planning algorithm to identify the shortest path that satisfies the two reliability constraints.

2. The global path planning method for an unmanned vehicle according to claim 1, characterized in that: The method of obtaining terrain information and soil information to characterize the terrain conditions of the target area includes the following steps: The target mapping is called the configuration space and is represented by Indicates; In the configuration space, the space with obstacles is called obstacle space, denoted by ; The space in which the unmanned vehicle can complete its mission is called free space. express; Any position in configuration space has two dimensions, given by given; The starting point is also called the initial point. Indicates that the end point is the goal to be achieved, with Represents; the connection between any nodes is called an edge; the planning path of the task is defined by Γ.

3. The global path planning method for an unmanned vehicle according to claim 2, characterized in that: Setting up the mobility model and path planning algorithm The RRT* algorithm includes the following steps, The maneuverability of a vehicle is closely related to vehicle dynamics and terrain conditions; the generalized formula for the maneuverability model is as follows: in It means the vehicle is in position The output of mobility, is a general mobility model, is the vector of vehicle related parameters, is a vector representing the conditions around the vehicle; The path planning algorithm RRT* is obtained by The iteration of the growing tree structure begins to find the optimal path; a random node is generated using a sampling strategy ;if , is rejected; the sampling process continues until there is a random sample in the free space, i.e. ; Then identify the tree T The nearest node ;if Can be Access, which means and There is no obstacle between By connecting and is inserted into the tree; Otherwise, use the redirect function to generate a new node and by connecting it to Add to the tree; Perform collision checks to ensure and There are no obstacles between them; After passing the collision check, the tree is modified using the rewiring operation; the above operations are performed iteratively until the optimal path is found; In the mobility model, the mobility model is combined with the path planning algorithm RRT* to identify the optimal path while satisfying the physical constraints.

4. The global path planning method for an unmanned vehicle according to claim 1, characterized in that: The multimodal reliability analysis of unmanned vehicles under uncertain terrain conditions includes: S3.

1. Sources of uncertainty and modeling The uncertainty comes from terrain and soil related parameters, and Gaussian random fields are used to model the uncertainty of the parameters; the spatial correlation between any two locations on the map is described by the autocorrelation function as follows: In the formula is the relevant length along the space; , are any two points in the graph, is the correlation between these two points; In order to generate random realizations of the random field for reliability analysis, the truncated Karhunen-Loeve expansion is adopted as follows: in is a sample of generated realizations of a random field; is the spatial coordinate of the random field represents the mean function; M is the cutoff number; and are the eigenvalues ​​and eigenvectors obtained by eigenanalysis of the correlation matrix; and are independent standard Gaussian random variables; S3.2 Reliability analysis of maximum achievable speed S3.2.1 The reliability of the maximum achievable speed is defined as follows: in is the maximum achievable speed, is the speed threshold, " stands for "for all", and is the spatial location The terrain and soil related parameter vectors are: ; S3.2.

2. Given the training data, the initial Gaussian process model GP is obtained in the high-fidelity simulation model as: in is the maximum achievable speed predicted by GP; and are the vectors representing the basis functions and the corresponding coefficients respectively; is the error term; In addition to the training data, a set of input samples is generated using the Monte Carlo simulation method MCS, which is iteratively improved on the GP model given above; in each iteration, the identified new training samples are added to the training set to improve the GP model; after the algorithm converges, the safe area and failure area of ​​the vehicle's maximum achievable speed are determined; MCS can be used to calculate the vehicle's mission maneuverability reliability expressed in terms of the maximum achievable speed: S3.3 Reliability analysis of vertical acceleration S 3.3.

1. The vertical acceleration is insensitive to soil parameters, and the effect of soil type on vertical acceleration is ignored; the vertical acceleration model is simplified as a function of terrain elevation as follows: in For spatial location The vertical acceleration at ( )for The elevation of and is a nonlinear function, and the model is also called the nonlinear autoregressive exogenous model NARX; S3.3.

2. First, the Gaussian mixture model GMM is used to segment multiple NARX models of different segments from the entire prediction area. Then, the NARX models are dynamically integrated on the path to predict the vertical acceleration along the path. The reliability of the vertical acceleration is as follows: in, is the acceleration predicted by the NARX model along the planned path; Vertical acceleration threshold.

5. The global path planning method for an unmanned vehicle according to claim 4, characterized in that: The RRT algorithm is combined with the agent model to explore the path and find the optimal path, including: S4.

1. Considering these two failure modes in unmanned vehicle path planning, the reliability-based path planning model is as follows: st In the formula To design the path, So To minimize the cost, and Mission maneuver reliability is expressed in terms of maximum achievable speed and vertical acceleration, respectively; is the reliability target in terms of maximum achievable speed, is the reliability target in terms of vertical acceleration; First, a tree represented by T is initialized, which contains all the nodes generated during the path exploration process; in the first step, there is only the starting point and goals In t; Next, randomly generate a node , and find the distance in the tree The nearest node ( ); it is noticed that the first random node, the closest node is ; The function used to find the nearest node is as follows: After finding a random node, use a steering function to generate a new node , expressed as: in represents the steering function; then arrive Reliability and Check; if the path meets the reliability target, find a neighbor node within the radius and record it as: The radius is defined as: in is a constant defined according to the environment, is the number of iterations, dim is the dimension of the configuration space; After that, at the lowest cost from Select The algorithm performs rewiring and rebuilds to The above operations are iterated until The area where the target point is reached.

6. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 5.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 5.

Images