Layered fast path planning method for large-scale complex environment
By employing a hierarchical fast path planning method, combined with depth-first search, improved RRT and RRT* algorithms, the problem of low path planning efficiency in large-scale complex environments is solved, generating high-precision, asymptotically optimal paths that adapt to dynamic environmental changes.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-28
- Publication Date
- 2026-04-14
AI Technical Summary
Existing path planning algorithms suffer from high computational complexity, numerous redundant exploration nodes, and poor path quality in large-scale complex environments, making it difficult to meet the requirements for high accuracy and rapid planning. In particular, they are inefficient in searching under tiled maps and incomplete information.
A hierarchical fast path planning method is adopted, which uses a depth-first search algorithm for long-range planning and combines an improved RRT and RRT* algorithm for local path planning. A heuristic exploration mechanism and historical path caching are added, and the sampling strategy is optimized to generate smooth and short paths.
It achieves high-precision and rapid generation of asymptotically optimal paths in large-scale complex environments, improving the efficiency and robustness of path planning, supporting dynamic replanning, and adapting to dynamic environmental changes.
Smart Images
Figure CN121855561A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of artificial intelligence technology and relates to a hierarchical fast path planning method for large-scale complex environments. Background Technology
[0002] In path planning on high-resolution tiled maps, traditional path planning algorithms such as Dijkstra and A* face severe challenges in terms of computational complexity and storage requirements due to the hierarchical structure and massive number of nodes. Traditional path planning methods tend to generate a large number of redundant exploration nodes on large-scale maps, and the computational complexity increases rapidly with the map's range and accuracy, resulting in low search efficiency. Furthermore, traditional algorithms typically assume that the map is static and complete; however, under incomplete information, the paths generated by traditional methods are of poor quality and fail to meet accuracy requirements. Heuristic planning methods may also become inapplicable under incomplete information. In contrast, existing technologies such as Rapid Exploratory Random Tree (RRT) and its improved algorithm RRT* can relatively effectively cope with environmental uncertainties and complexities in high-precision, incomplete information path planning, providing asymptotically optimal path planning solutions, thus demonstrating certain advantages in practical applications.
[0003] The random sampling method used by the RRT algorithm to construct the search tree can easily generate a large number of redundant exploration nodes on large-scale maps, resulting in low search efficiency and difficulty in achieving the smoothness and brevity required for high-precision planning. Although the RRT* algorithm can optimize the path, in complex search spaces composed of a large number of irregular obstacles, without the guidance of heuristic functions, the computational load is large and the planning time is too long. In addition, existing algorithms lack effective planning strategies for tiled map structures and high-precision coordinate systems, making it difficult to achieve fast path planning while ensuring planning accuracy. Summary of the Invention
[0004] The technical problem solved by this invention is to overcome the shortcomings of the prior art and propose a hierarchical fast path planning method for large-scale complex environments. This method addresses the problems of low planning efficiency, poor path quality, search space explosion, and the significant impact of frequent replanning in large-scale high-precision path planning with incomplete information, caused by tiled maps and the requirement for high-precision coordinates below the meter level.
[0005] The solution to the technical problem of this invention is: a hierarchical fast path planning method for large-scale complex environments, comprising the following steps:
[0006] Step 1: Based on the connectivity characteristics of the map region, segment the tile map to construct a topology graph. Treat each sub-region as a node and use the depth-first search algorithm for long-range planning to obtain the approximate direction of movement from the starting point to the target point and the critical path nodes.
[0007] Step 2: Within each sub-region determined by the depth-first search algorithm, perform adaptive local path planning: Improve the sampling strategy of the RRT and RRT* algorithms by adding a heuristic exploration mechanism during the sampling process. When planning the path, first call the improved RRT algorithm to sample the surrounding environment and identify the complexity of the obstacle environment. For sub-regions with relatively simple obstacle environments, directly use the improved RRT algorithm to quickly generate a feasible path. For sub-regions with relatively complex obstacle environments, switch to call the improved RRT* algorithm for planning to obtain an asymptotically optimal path.
[0008] Step 3: Cache historical planning paths. When the environment changes and the path needs to be replanned, match the critical path nodes in the current environment with those in the historical environment. If the match is successful, check whether the corresponding historical planning path is reachable in the current environment. If it is reachable, integrate the historical planning path into the new planning task. Otherwise, repeat step 2.
[0009] Furthermore, the improved RRT algorithm has the following process:
[0010] 1) Initialize, obtain the start and end points, and set the number of iterations;
[0011] 2) Execute the following loop according to the number of iterations:
[0012] The step size is adaptively adjusted based on the distance from the current position to the destination and the number of existing nodes.
[0013] Random sampling node q;
[0014] Neighbor retrieval: Retrieve the nearest neighbor qNearest of node q from the existing nodes in the current spanning tree structure; if there is no node in the current spanning tree structure, then node q is taken as the first node;
[0015] A new node, newNode, is generated through a local exploration strategy: the local exploration process of newNode involves sampling at equal intervals within the angular constraints between the exploration direction and the target direction, and adjusting the direction and distance of the final sampling point according to the proportion of unreachable points; the exploration direction is given by a heuristic exploration mechanism, and the target direction is the direction from qNearest to the endpoint;
[0016] Check if the path from qNearest to the new node newNode is unobstructed. If it is unobstructed, insert the new node newNode and update the spanning tree structure. Otherwise, proceed to the next iteration.
[0017] If the endpoint is reached or the number of explored nodes exceeds the limit, exit the loop;
[0018] 3) Path extraction: Backtrack from the last node generated in the spanning tree to obtain the search path.
[0019] Furthermore, the improved RRT* algorithm has the following process:
[0020] 1) Initialize, obtain the start and end points, and set the number of iterations;
[0021] 2) Execute the following loop according to the number of iterations:
[0022] The step size is adaptively adjusted based on the distance from the current position to the destination and the number of existing nodes.
[0023] Random sampling node q;
[0024] Neighbor retrieval: Retrieve the nearest neighbor group qNearest of node q from the existing nodes in the current spanning tree structure; if there are no nodes in the current spanning tree structure, then node q is taken as the first node;
[0025] A new node, newNode, is generated through a local exploration strategy: the local exploration process of newNode involves sampling at equal intervals within the angular constraints between the exploration direction and the target direction, and adjusting the direction and distance of the final sampling point according to the proportion of unreachable points; the exploration direction is given by a heuristic exploration mechanism, and the target direction is the direction from qNearest to the endpoint;
[0026] To select the optimal parent node of the new node `newNode`, minimize the path cost from the starting point to the new node `newNode`: perform reconnection, attempting to connect each neighboring point in the point group `qNearest` to the new node `newNode`, and calculate the new cost from the new node `newNode` to the starting point = the cost from the new node `newNode` to the neighboring point + the cost from the neighboring point to the starting point. The cost from the neighboring point to the starting point has already been calculated in previous loops. When there is an obstacle between the new node `newNode` and its neighboring points, assign the cost from the new node `newNode` to the neighboring point a preset maximum value. If the new cost is smaller than the original cost, update the parent node of the new node `newNode` to the neighboring point.
[0027] Insert a new node (newNode) and update the spanning tree structure;
[0028] If the endpoint is reached or the number of explored nodes exceeds the limit, exit the loop;
[0029] 3) Path extraction: Backtrack from the last node generated in the spanning tree to obtain the search path.
[0030] Furthermore, the relatively simple obstacle environment sub-regions include the flat roadside area.
[0031] Furthermore, the sub-region with relatively complex obstacle environment refers to an area containing a large number of irregular and inaccessible locations, including residential areas with buildings, mountainous and hilly areas, ponds and rivers.
[0032] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of a hierarchical fast path planning method for large-scale complex environments.
[0033] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of a hierarchical fast path planning method for large-scale complex environments.
[0034] A computer program product includes a computer program that, when executed by a processor, implements the steps of a hierarchical fast path planning method for large-scale complex environments.
[0035] The advantages of this invention compared to the prior art are:
[0036] (1) This invention has online high-precision planning capability. Through hierarchical local planning, RRT* rewiring and heuristic exploration strategy, it obtains better sampling points, generates smooth and short paths, has robust path error control, supports dynamic replanning, and improves the adaptability and stability of the agent in dynamic environments.
[0037] (2) The present invention can quickly generate asymptotically optimal paths using limited computing resources under large-scale complex incomplete maps, thereby achieving rapid completion of long-range planning of the whole map. Attached Figure Description
[0038] Figure 1 This is a flowchart of a hierarchical rapid path planning method for large-scale complex environments according to the present invention.
[0039] Figure 2 This is a schematic diagram of the long-range planning of the DFS algorithm in an embodiment of the present invention;
[0040] Figure 3 This is a schematic diagram illustrating the use of the improved RRT algorithm in a simple area according to an embodiment of the present invention;
[0041] Figure 4 This is a schematic diagram illustrating the use of the improved RRT* algorithm in complex regions according to an embodiment of the present invention. Detailed Implementation
[0042] like Figure 1As shown, the present invention proposes a hierarchical fast path planning method for large-scale complex environments, comprising the following steps:
[0043] Step 1: Based on the connectivity characteristics of the map region, segment the tile map to construct a topology graph. Treat each sub-region as a node and use the depth-first search (DFS) algorithm for long-range planning to obtain the approximate direction of movement from the starting point to the target point and the critical path nodes.
[0044] Step 2: Within each sub-region determined by the Depth-First Search (DFS) algorithm, perform adaptive local path planning: Improve the sampling strategy of the RRT and RRT* algorithms by incorporating a heuristic exploration mechanism during the sampling process. When planning the path, first call the improved RRT algorithm to sample the surrounding environment and identify the complexity of the obstacle environment. For sub-regions with relatively simple obstacle environments, directly use the improved RRT algorithm to quickly generate a feasible path. For sub-regions with relatively complex obstacle environments, switch to calling the improved RRT* algorithm for planning to obtain an asymptotically optimal path.
[0045] Step 3: Cache historical planning paths. When the environment changes and the path needs to be replanned, match the critical path nodes in the current environment with those in the historical environment. If the match is successful, check whether the corresponding historical planning path is reachable in the current environment. If it is reachable, integrate the historical planning path into the new planning task. Otherwise, repeat step 2.
[0046] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0047] Example 1
[0048] like Figure 1 , Figure 2 As shown, the hierarchical fast path planning method in this embodiment has the following steps:
[0049] Step 1: Based on the connectivity characteristics of the map region, segment the tile-based map to construct a topological graph. Treat each segmented sub-region as a node and use the Depth-First Search (DFS) algorithm for long-range planning to obtain the approximate direction of movement from the starting point to the target point and the critical path nodes. For example, bridge nodes connecting the two banks of a river.
[0050] Step 2: Within each sub-region determined by the Depth-First Search (DFS) algorithm, perform adaptive local path planning: Improve the sampling strategy of the RRT and RRT* algorithms by incorporating a heuristic exploration mechanism during the sampling process. When planning the path, first call the improved RRT algorithm to sample the surrounding environment and identify the complexity of the obstacle environment. For sub-regions with relatively simple obstacle environments, directly use the improved RRT algorithm to quickly generate a feasible path. For sub-regions with relatively complex obstacle environments, switch to calling the improved RRT* algorithm for planning to obtain an asymptotically optimal path.
[0051] The sampling strategy for the RRT and RRT* algorithms is improved by incorporating a heuristic exploration mechanism during the sampling process. The idea is as follows:
[0052] Improved RRT / RRT* global scheduling method:
[0053] 1. Initialize the PathPlanner:
[0054] Input: Agent model, startPos, endPos
[0055] Initialize the path planner
[0056] 2. Configure the path planner parameters:
[0057] Define two path containers:
[0058] total_path_rrt_star: Stores the planned path for RRT*.
[0059] total_path_rrt: Stores alternative paths for RRT.
[0060] 3. Calculate the distance between the starting point and the ending point:
[0061] real_dis=pathPlanner.myLib.Distance(startPos,endPos)
[0062] 4. Configure RRT and RRT* parameters:
[0063] pathPlanner.setStepSize(stepSize)
[0064] pathPlanner.setStepNum()
[0065] pathPlanner.setMaxIterations() / / Number of samples
[0066] pathPlanner.setRandomSize() / / Sets the sampling probability at the endpoint
[0067] 5. Perform RRT route planning:
[0068] total_path_rrt=pathPlanner.doPathPlanningWithTime(startTime)
[0069] 6. Determine if RRT was successful:
[0070] If RRT fails or the number of nodes exceeds the limit:
[0071] Adjust the parameters and replan using RRT* mode:
[0072] total_path_rrt_star=pathPlanner.doPathPlanningWithTime(startTime,1)
[0073] 7. Select the final path:
[0074] If RRT* succeeds:
[0075] total_path=total_path_rrt_star
[0076] elif RRT* fails but RRT succeeds:
[0077] total_path = total_path_rrt
[0078] else:
[0079] Neither method found the complete path.
[0080] Return area unreachable
[0081] Based on the above design ideas, such as Figure 3 As shown, the improved RRT algorithm flow is as follows:
[0082] 1) Initialize, obtain the start and end points, and set the number of iterations;
[0083] 2) Execute the following loop according to the number of iterations:
[0084] The step size is adaptively adjusted based on the distance from the current position to the destination and the number of existing nodes.
[0085] Random sampling node q;
[0086] Neighbor retrieval: Retrieve the nearest neighbor qNearest of node q from the existing nodes in the current spanning tree structure; if there is no node in the current spanning tree structure, then node q is taken as the first node;
[0087] A new node, newNode, is generated through a local exploration strategy: the local exploration process of newNode involves sampling at equal intervals within the angular constraints between the exploration direction and the target direction, and adjusting the direction and distance of the final sampling point according to the proportion of unreachable points; the exploration direction is given by a heuristic exploration mechanism, and the target direction is the direction from qNearest to the endpoint;
[0088] Check if the path from qNearest to the new node newNode is unobstructed. If it is unobstructed, insert the new node newNode and update the spanning tree structure. Otherwise, proceed to the next iteration.
[0089] If the endpoint is reached or the number of explored nodes exceeds the limit, exit the loop;
[0090] 3) Path extraction: Backtrack from the last generated node in the spanning tree to obtain the search path; call the NodeToLine function to convert the path node into a line segment and return the converted path.
[0091] Based on the above design ideas, such as Figure 4 As shown, the improved RRT* algorithm flow is as follows:
[0092] 1) Initialize, obtain the start and end points, and set the number of iterations;
[0093] 2) Execute the following loop according to the number of iterations:
[0094] The step size is adaptively adjusted based on the distance from the current position to the destination and the number of existing nodes.
[0095] Random sampling node q;
[0096] Neighbor retrieval: Retrieve the nearest neighbor group qNearest of node q from the existing nodes in the current spanning tree structure; if there are no nodes in the current spanning tree structure, then node q is taken as the first node;
[0097] A new node, newNode, is generated through a local exploration strategy: the local exploration process of newNode involves sampling at equal intervals within the angular constraints between the exploration direction and the target direction, and adjusting the direction and distance of the final sampling point according to the proportion of unreachable points; the exploration direction is given by a heuristic exploration mechanism, and the target direction is the direction from qNearest to the endpoint;
[0098] To select the optimal parent node of the new node `newNode`, minimize the path cost from the starting point to the new node `newNode`: perform reconnection, attempting to connect each neighboring point in the point group `qNearest` to the new node `newNode`, and calculate the new cost from the new node `newNode` to the starting point = the cost from the new node `newNode` to the neighboring point + the cost from the neighboring point to the starting point. The cost from the neighboring point to the starting point has already been calculated in previous loops. When there is an obstacle between the new node `newNode` and its neighboring points, assign the cost from the new node `newNode` to the neighboring point a preset maximum value. If the new cost is smaller than the original cost, update the parent node of the new node `newNode` to the neighboring point.
[0099] Insert a new node (newNode) and update the spanning tree structure;
[0100] If the endpoint is reached or the number of explored nodes exceeds the limit, exit the loop;
[0101] 3) Path extraction: Backtrack from the last generated node in the spanning tree to obtain the search path; call the NodeToLine function to convert the path node into a line segment and return the converted path.
[0102] The sub-region with a relatively simple obstacle environment mentioned in Step 2 refers to the flat area along the highway. The sub-region with a relatively complex obstacle environment refers to residential areas containing buildings, mountainous and hilly areas, ponds and rivers, and other areas containing a large number of irregular and inaccessible locations.
[0103] Step two addresses the search space explosion problem caused by massive nodes and irregular obstacles by improving the sampling strategies of the RRT and RRT* algorithms. For accuracy optimization, a local heuristic exploration mechanism is used to expand the exploration around the sampling point, focusing more on points on the highway network and avoiding path deviations and unnecessary detours. The reconnection and local adjustment methods of RRT* are used to optimize the sampling path, selecting easily passable areas to ensure asymptotically optimal paths. For speed improvement, a heuristic exploration mechanism is added during the sampling process to achieve variable step size sampling, dynamically adjusting the search step size based on the target distance and surrounding environmental characteristics. When the target distance is close or the surrounding environment is simple, the search step size is increased to quickly traverse obstacle-free areas; when approaching the target or in complex and irregular areas, the search step size is decreased for refined sampling planning, achieving fast and high-precision sampling planning in special irregular areas.
[0104] Step 3: Cache historical planning paths. When the environment changes and the path needs to be replanned, match the critical path nodes in the current environment with those in the historical environment. If the match is successful, check whether the corresponding historical planning path is reachable in the current environment. If it is reachable, integrate the historical planning path into the new planning task. Otherwise, repeat step 2.
[0105] In summary, this invention achieves high-precision and efficient path planning for hierarchical agents in large-scale complex environments by using innovative graph layering and hierarchical planning methods, combined with DFS long-range planning, RRT* and RRT local planning, as well as heuristic search, improved sampling strategies and historical planning caching mechanisms, thus meeting the stringent requirements for planning accuracy and real-time performance in practical applications.
[0106] This application provides a computer-readable storage medium storing computer instructions that, when executed on a computer, cause the computer to perform... Figure 1 The method described.
[0107] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage and optical storage) containing computer-usable program code.
[0108] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0109] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0110] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0111] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.
[0112] The contents not described in detail in this specification are common knowledge to those skilled in the art.
Claims
1. A hierarchical fast path planning method for large-scale complex environments, characterized in that, Includes the following steps: Step 1: Based on the connectivity characteristics of the map region, segment the tile map to construct a topology graph. Treat each sub-region as a node and use the depth-first search algorithm for long-range planning to obtain the approximate direction of movement from the starting point to the target point and the critical path nodes. Step 2: Within each sub-region determined by the depth-first search algorithm, perform adaptive local path planning: Improve the sampling strategy of the RRT and RRT* algorithms by adding a heuristic exploration mechanism during the sampling process. When planning the path, first call the improved RRT algorithm to sample the surrounding environment and identify the complexity of the obstacle environment. For sub-regions with relatively simple obstacle environments, directly use the improved RRT algorithm to quickly generate a feasible path. For sub-regions with relatively complex obstacle environments, switch to call the improved RRT* algorithm for planning to obtain an asymptotically optimal path. Step 3: Cache historical planning paths. When the environment changes and the path needs to be replanned, match the critical path nodes in the current environment with those in the historical environment. If the match is successful, check whether the corresponding historical planning path is reachable in the current environment. If it is reachable, integrate the historical planning path into the new planning task. Otherwise, repeat step 2.
2. The hierarchical rapid path planning method for large-scale complex environments according to claim 1, characterized in that, The improved RRT algorithm has the following process: 1) Initialize, obtain the start and end points, and set the number of iterations; 2) Execute the following loop according to the number of iterations: The step size is adaptively adjusted based on the distance from the current position to the destination and the number of existing nodes. Random sampling node q; Neighbor retrieval: Retrieve the nearest neighbor qNearest of node q from the existing nodes in the current spanning tree structure; if there is no node in the current spanning tree structure, then node q is taken as the first node; A new node, newNode, is generated through a local exploration strategy: the local exploration process of newNode involves sampling at equal intervals within the angular constraints between the exploration direction and the target direction, and adjusting the direction and distance of the final sampling point according to the proportion of unreachable points; the exploration direction is given by a heuristic exploration mechanism, and the target direction is the direction from qNearest to the endpoint; Check if the path from qNearest to the new node newNode is unobstructed. If it is unobstructed, insert the new node newNode and update the spanning tree structure. Otherwise, proceed to the next iteration. If the endpoint is reached or the number of explored nodes exceeds the limit, exit the loop; 3) Path extraction: Backtrack from the last node generated in the spanning tree to obtain the search path.
3. The hierarchical rapid path planning method for large-scale complex environments according to claim 1, characterized in that, The improved RRT* algorithm has the following process: 1) Initialize, obtain the start and end points, and set the number of iterations; 2) Execute the following loop according to the number of iterations: The step size is adaptively adjusted based on the distance from the current position to the destination and the number of existing nodes. Random sampling node q; Neighbor retrieval: Retrieve the nearest neighbor group qNearest of node q from the existing nodes in the current spanning tree structure; if there are no nodes in the current spanning tree structure, then node q is taken as the first node; A new node, newNode, is generated through a local exploration strategy: the local exploration process of newNode involves sampling at equal intervals within the angular constraints between the exploration direction and the target direction, and adjusting the direction and distance of the final sampling point according to the proportion of unreachable points; the exploration direction is given by a heuristic exploration mechanism, and the target direction is the direction from qNearest to the endpoint; To select the optimal parent node of the new node `newNode`, minimize the path cost from the starting point to the new node `newNode`: perform reconnection, attempting to connect each neighboring point in the point group `qNearest` to the new node `newNode`, and calculate the new cost from the new node `newNode` to the starting point = the cost from the new node `newNode` to the neighboring point + the cost from the neighboring point to the starting point. The cost from the neighboring point to the starting point has already been calculated in previous loops. When there is an obstacle between the new node `newNode` and its neighboring points, assign the cost from the new node `newNode` to the neighboring point a preset maximum value. If the new cost is smaller than the original cost, update the parent node of the new node `newNode` to the neighboring point. Insert a new node (newNode) and update the spanning tree structure; If the endpoint is reached or the number of explored nodes exceeds the limit, exit the loop; 3) Path extraction: Backtrack from the last node generated in the spanning tree to obtain the search path.
4. The hierarchical rapid path planning method for large-scale complex environments according to claim 1, characterized in that, The sub-regions with relatively simple obstacle environments include the flat areas along highways.
5. The hierarchical rapid path planning method for large-scale complex environments according to claim 1, characterized in that, The sub-region with relatively complex obstacle environment refers to an area containing a large number of irregular and inaccessible locations, including residential areas with buildings, mountainous and hilly areas, ponds and rivers.
6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 5.
7. 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, it implements the steps of the method as described in any one of claims 1 to 5.
8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 5.