Unmanned aerial vehicle coverage path planning method for urban area

Through the adaptive drone flight altitude model and area decomposition method, complex urban areas are decomposed into feasible subspaces and transformed the problem into a generalized travel dealer problem, solving the problems of high computational complexity and many optimization variables in the existing technology. , realize efficient and secure drone coverage path planning within urban areas.

CN119043340AActive Publication Date: 2024-11-29HUNAN UNIV

Patent Information

Application Number
CN202411524162.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-30
Publication Date
2024-11-29
Estimated Expiration
2044-10-30

AI Technical Summary

Technical Problem

The existing UAV coverage path planning method has problems such as high computational complexity, many optimization variables, and insufficient planning optimization when dealing with complex urban areas, which is difficult to meet traffic monitoring, network services, air patrol, fire monitoring and emergency rescue. Application requirements.

Method used

Adaptive drone flight altitude model and area decomposition method are used to decompose the target area into multiple feasible subspaces without obstacles, and the optimization problem is equivalent to a generalized travel dealer problem, by designing the multi-approximate condition Objective function, quickly solve the overlay path.

Benefits of technology

It realizes efficient, safe and flexible drone coverage path planning in urban areas, reduces the computational complexity and number of subspaces, and improves the coverage efficiency and optimal planning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119043340A_ABST
    Figure CN119043340A_ABST
Patent Text Reader

Abstract

The invention discloses an unmanned aerial vehicle coverage path planning method for an urban area, and the method comprises the steps: determining the starting point and terminal point coordinates of an unmanned aerial vehicle coverage task, and building a map model according to a terrain database of a region where the flight task is located; a height model of unmanned aerial vehicle coverage flight is established, the grid map is preprocessed according to the model, and a feasible space of a coverage task is obtained; designing an integer programming method of set division, decomposing the target area into a plurality of feasible subspaces without obstacles, and obtaining a space decomposition graph of the task area; based on the space decomposition graph, constructing an optimization problem which takes the minimum path length as the cost and takes the starting point, the ending point and the coverage direction of the subspace coverage planning as the output; and equivalently converting the optimization problem into a generalized traveling salesman problem for solving, generating a path covering each subspace, and completing path planning covering the whole target area. The method has the advantages of high flexibility, low calculation complexity, high robustness, high safety and the like.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) path planning, and in particular relates to a UAV coverage path planning method for urban areas. Background Art

[0002] Drones have been widely used due to their strong maneuverability and high flexibility. Especially in urban areas, drones play an important role in traffic monitoring, network services, aerial patrols, fire monitoring and emergency rescue. Therefore, it is extremely important to design a safe, efficient and feasible drone coverage path planning method for urban areas, which is of great significance to improving the efficiency, safety and reliability of drone operations in cities. The goal of coverage path planning is to plan a series of paths under the premise of meeting the constraints of drone movement and mission space, so that the drone can fully cover the target area and ensure the optimal time and energy consumption indicators during the flight.

[0003] Traditional UAV coverage path planning mostly uses the grid method. First, the target area is divided into a series of closely connected and uniformly sized grids. Then, the UAV visits all grids in turn to achieve coverage of the entire target area. However, as the area of ​​the target area increases, the computational memory occupied by the grid increases linearly, resulting in an exponential increase in the computational complexity of solving feasible coverage paths. Therefore, the grid method is difficult to apply to large area coverage tasks. Unlike the grid method, by designing a series of parallel UAV reciprocating motion paths, a large area can be efficiently covered without complex calculations. However, the reciprocating motion path cannot handle obstacles in the area and can mostly only be applied to coverage tasks in open areas, such as UAV spraying pesticides. In order to solve the coverage planning problem in the obstacle space, researchers began to study the coverage planning method based on graph theory. The core idea is to obtain multiple feasible subspaces by segmenting the target area, and then construct a path optimization algorithm on this basis to achieve the optimal planning of the coverage path. Compared with other methods, it solves the coverage planning results by traversing the feasible subspace, which effectively reduces the computational complexity, and the subspace segmentation can significantly reduce the number of turns and flight distance of the UAV, thereby improving the coverage efficiency. However, the existing segmentation methods do not fully merge subspaces, and there are bottleneck problems such as multiple subspaces, complex optimization variables, slow solution speed, and insufficient planning optimality. In addition, the coverage planning efficiency and spatial constraints are not fully considered in the construction of the optimization problem, resulting in the coverage planning results being difficult to apply to complex urban areas.

[0004] In order to further promote the application of drones in traffic monitoring, network services, aerial patrols, fire monitoring and emergency rescue in complex urban areas, it is urgent to invent an efficient and safe drone coverage path planning method. Summary of the invention

[0005] In view of the above technical problems, the present invention provides a UAV coverage path planning method for urban areas.

[0006] The technical solution adopted by the present invention to solve the technical problem is: A method for planning a UAV coverage path for an urban area, the method comprising the following steps: S100: Determine the starting point and end point coordinates of the UAV coverage mission, and build a map model based on the terrain database of the area where the flight mission is located; S200: Establishing a height model for the UAV coverage flight, preprocessing the grid map according to the height model, and obtaining a feasible space for the coverage task; S300: Design an integer programming method for set partitioning to decompose the target area into multiple feasible subspaces that do not contain obstacles, and obtain a spatial decomposition diagram of the task area; S400: Based on the space decomposition graph, construct an optimization problem with the cost of minimizing the path length and the starting point, end point and covering direction of the subspace coverage planning as output; S500: The optimization problem is equivalently converted into a generalized traveling salesman problem for solution, a path covering each subspace is generated, and path planning covering the entire target area is completed.

[0007] Preferably, S100 includes: S110: Determine the starting point and end point coordinates of the UAV flight according to the UAV coverage mission requirements and initial conditions; S120: Generate a digital topographic map of the target area using the terrain data, clearly define the horizontal and vertical boundaries of the coverage task, and the vertex positions and height information of obstacles in the area; S130: Convert digitized topographic map into raster map: ; in Represents the target area, Represents the horizontal and vertical coordinates of the drone, Represents the height of the grid point, represents the entire set of real numbers.

[0008] Preferably, S200 includes: S210: Establish UAV altitude model; S220: Before executing the mission, adjust the altitude of the drone's coverage flight according to the mission requirements and the target area environmental information, and pre-process the raster map of the area according to the flight altitude, wherein the obstacle area greater than the drone's flight altitude is regarded as an inaccessible area, and the obstacle area less than the drone's flight altitude is regarded as a reachable area, and generate a target area raster map for coverage planning.

[0009] Preferably, S210 specifically includes: ; in Represents the altitude of the drone, Represents the requirement for clarity, represents the clarity weight, Represents the wind factor in the target area, represents the safety weight, It is the standard flight altitude of drones; The specific processing process in S220 is as follows: ; in represents the flight altitude of the drone, and k represents the altitude of the drone. The accessibility of Indicates that the drone is unreachable. Represents drone reachable; The new grid map is as follows: ; According to the query The value of can determine the corresponding Obstacle situation at the location.

[0010] Preferably, S300 includes: S310: Extending edges of obstacles with inner angles less than 180 degrees until they hit a map boundary or other obstacles, and extending the edges to divide the target space into multiple subspaces; S320: merging the divided subspaces by using an integer programming algorithm for set partitioning, thereby reducing the number of subspaces; S330: Decomposing the target area into a plurality of sub-areas that do not contain obstacles by solving an integer programming problem.

[0011] Preferably, the integer programming model for set partitioning in S320 is specifically: ; ; ; ; in By strategy The width of the generated convex polygon, represents the selected strategy, Represents all strategies that can be merged, , Indicates whether to select a strategy , 1 if selected, 0 otherwise, represents the subspace, Represents a collection of subspaces. If the merging strategy Include Sub-Polygons ,but is 1, otherwise it is 0; Preferably, S400 includes: S410: Based on the obtained task area spatial decomposition diagram, it is necessary to determine the starting point, end point and covering direction of the subspace, wherein the direction selection depends on the direction of each edge of the subspace, that is: ,in Represents the covering direction, F represents all the edges of the subspace, and the starting and ending points are the vertices of each subspace, that is: ,in , Represent the starting point and the end point respectively, and V represents all the nodes in the subspace; S420: Establishing an optimization problem with minimizing the path length as the cost and taking the starting point, end point and coverage direction of the subspace coverage planning as the output; Preferably, the optimization problem of the path cost in S420 is specifically: ; in, is the total number of sub-regions decomposed, ) is the The sub-regions cover the coordinates of the path end points, ) is the The sub-regions cover the coordinates of the path starting point, Represents the cost of transitioning between different subspaces.

[0012] Preferably, S500 includes: S510: Define adjacency graph: ,in represents the set of subspaces, Represents the access order, that is, from arrive The order in which access is performed; each access sequence There is a arrive The shortest path , use the existing graph search algorithm to obtain the distance of the shortest path; S520: The path is planned using the speed ramp function parameterization. The total cost of the trajectory is the sum of the costs of all segmented trajectories. The cost of each segment of the trajectory is The value depends on the distance between the two points 、Maximum speed of drone and maximum acceleration : ; in , , is the time to accelerate to maximum speed, is the distance the drone travels when accelerating, is the maximum speed, is the maximum acceleration, For arrive The shortest path; since the starting and ending points of the area coverage do not coincide, the cost matrix is asymmetric, that is ; S530: Based on S300, the vertex set of each subspace is fixed. By selecting appropriate starting and ending points and the order of visiting the subspaces to ensure the minimization of the transition time, the following generalized traveling salesman problem is constructed: ; in Represents the total cost, The first visit represents the starting point. The last visited one is the end point. Indicates that all nodes have been visited; The computational geometry algorithm library CGAL is used to solve the generalized traveling salesman problem, and the final path obtained can achieve optimal coverage of the urban area.

[0013] The above-mentioned UAV coverage path planning method for urban areas proposes an adaptive UAV flight altitude model to enable the UAV to dynamically adjust the flight altitude under uncertain environments, thereby meeting different precision task requirements; a regional decomposition method is proposed to deal with obstacles in urban areas and reduce the number of sub-target areas; the problem is equivalently converted into a generalized traveling salesman problem, and by designing an objective function under multiple constraints, it can be solved quickly. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flow chart of a method for planning a UAV coverage path for an urban area in one embodiment of the present invention; Figure 2 It is a schematic diagram of the initial map of the present invention; Figure 3 A pre-processed map of the present invention, including a starting point, an end point and obstacles; Figure 4 This is a schematic diagram of the present invention after regional decomposition, where the entire map is divided into several subspaces; Figure 5The present invention is directed to the selection of the subspace area coverage direction; Figure 6 The coverage result generated by the present invention for a given map; Figure 7 This is the path finally generated by the present invention. DETAILED DESCRIPTION

[0015] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings.

[0016] In one embodiment, Figure 1 As shown, a method for planning a UAV coverage path for an urban area comprises the following steps: S100: Determine the starting point and end point coordinates of the UAV coverage mission, and build a map model based on the terrain database of the area where the flight mission is located; S200: Establishing a height model for the UAV coverage flight, preprocessing the grid map according to the height model, and obtaining a feasible space for the coverage task; S300: Design an integer programming method for set partitioning to decompose the target area into multiple feasible subspaces that do not contain obstacles, and obtain a spatial decomposition diagram of the task area; S400: Based on the space decomposition graph, construct an optimization problem with the cost of minimizing the path length and the starting point, end point and covering direction of the subspace coverage planning as output; S500: The optimization problem is equivalently converted into a generalized traveling salesman problem for solution, a path covering each subspace is generated, and path planning covering the entire target area is completed.

[0017] The above-mentioned UAV coverage path planning method for urban areas proposes an adaptive UAV flight altitude model to enable the UAV to dynamically adjust the flight altitude under uncertain environments, thereby meeting different precision task requirements; a regional decomposition method is proposed to deal with obstacles in urban areas and reduce the number of sub-target areas; the problem is equivalently converted into a generalized traveling salesman problem, and by designing an objective function under multiple constraints, it can be solved quickly.

[0018] In one embodiment, S100 includes: S110: Determine the starting point and end point coordinates of the UAV flight according to the UAV coverage mission requirements and initial conditions; S120: Generate a digital topographic map of the target area using the terrain data, clearly define the horizontal and vertical boundaries of the coverage task, and the vertex positions and height information of obstacles in the area; S130: Convert digitized topographic map into raster map: ; in Represents the target area, Represents the horizontal and vertical coordinates of the drone, Represents the height of the grid point, represents the entire set of real numbers.

[0019] In one embodiment, S200 includes: S210: Establish UAV altitude model; S220: Before executing the mission, adjust the altitude of the drone's coverage flight according to the mission requirements and the target area environmental information, and pre-process the raster map of the area according to the flight altitude, wherein the obstacle area greater than the drone's flight altitude is regarded as an inaccessible area, and the obstacle area less than the drone's flight altitude is regarded as a reachable area, and generate a target area raster map for coverage planning.

[0020] Specifically, by comprehensively considering the wind conditions, visibility and other information in the city, a suitable altitude for covering the flight is generated. Figure 2 As shown in Figure 1, after selecting the flight altitude of the drone, there are obstacles whose height exceeds the flight plane of the drone. The preprocessed map is as follows Figure 3 shown.

[0021] In one embodiment, S210 specifically includes: ; in Represents the altitude of the drone, Represents the requirement for visibility, represents the visibility weight, Represents the wind factor in the target area, represents the safety weight, It is the standard flight altitude of drones; Take the DJI Mavic3E mapping drone as an example, its standard flight altitude When the visibility is 2 and the wind factor is 3, the design parameters are as required. They are -0.5 and -1 respectively. The flight altitude is 76.

[0022] The specific processing process in S220 is as follows: ; in represents the flight altitude of the drone, and k represents the altitude of the drone. The accessibility of Indicates that the drone is unreachable. Represents drone reachable; The new grid map is as follows: ; According to the query The value of can determine the corresponding Obstacle situation at the location.

[0023] In one embodiment, Figure 4 As shown, S300 includes: S310: Select an edge less than 180° in the obstacle space and extend it to divide the target space into multiple subspaces; S320: merging the divided subspaces by using an integer programming algorithm for set partitioning, thereby reducing the number of subspaces; S330: Decomposing the target area into a plurality of sub-areas that do not contain obstacles by solving an integer programming problem.

[0024] In one embodiment, the integer programming model for set partitioning in S320 is specifically: ; ; ; ; in By strategy The width of the generated convex polygon, represents the selected strategy, Represents all strategies that can be merged, , Indicates whether to select a strategy , 1 if selected, 0 otherwise, represents the subspace, Represents a collection of subspaces. If the merging strategy Include Sub-Polygons ,but is 1 if the value is true, otherwise it is 0.

[0025] Specifically, in the integer programming model of set partitioning, the first one is the objective function, and the second and third ones are constraints. The objective function minimizes the total width of the sub-polygons, the second constraint ensures that all initial sub-polygons are included in the selected merging scheme, and the third constraint is an integral constraint on the decision variables.

[0026] In one embodiment, Figure 5 As shown, S400 includes: S410: Based on the obtained task area spatial decomposition diagram, it is necessary to determine the starting point, end point and covering direction of the subspace, wherein the direction selection depends on the direction of each edge of the subspace, that is: ,in Represents the covering direction, F represents all the edges of the subspace, and the starting and ending points are the vertices of each subspace, that is: ,in , Represent the starting point and the end point respectively, and V represents all the nodes in the subspace; S420: Establish an optimization problem with minimizing the path length as the cost and taking the starting point, end point and coverage direction of the subspace coverage planning as the output.

[0027] Specifically, for a single subspace, the coverage direction with the least number of turns is the optimal solution, but for the global situation, the distance between the end point of the previous subspace and the starting point of the next subspace also needs to be considered. For the same coverage path, the nature of the path will change by choosing different path starting points and path endings.

[0028] In one embodiment, the optimization problem of the path cost in S420 is specifically: ; in, is the total number of sub-regions decomposed, ) is the The sub-regions cover the coordinates of the path end points, ) is the The sub-regions cover the coordinates of the path starting point, Represents the cost of transitioning between different subspaces.

[0029] Specifically, the planning problem of full coverage of the entire target area is transformed into an optimization problem of minimizing the path cost by selecting the starting point, end point and sequence of the subspace coverage path. In order to solve this problem efficiently, it is equivalently transformed into a generalized traveling salesman problem.

[0030] In one embodiment, S500 includes: S510: Define adjacency graph: ,in represents the set of subspaces, Represents the access order, that is, from arrive The order in which access is performed; each access sequence There is a arrive The shortest path , use the existing graph search algorithm to obtain the distance of the shortest path; S520: The path is planned using the speed ramp function parameterization. The total cost of the trajectory is the sum of the costs of all segmented trajectories. The cost of each segment of the trajectory is The value depends on the distance between the two points 、Maximum speed of drone and maximum acceleration : ; in , , is the time to accelerate to maximum speed, is the distance the drone travels when accelerating, is the maximum speed, is the maximum acceleration, For arrive The shortest path; since the starting and ending points of the area coverage do not coincide, the cost matrix is asymmetric, that is ; S530: Based on S300, the vertex set of each subspace is fixed. By selecting appropriate starting and ending points and the order of visiting the subspaces to ensure the minimization of the transition time, the following generalized traveling salesman problem is constructed: ; in Represents the total cost, The first visit represents the starting point. The last visited one is the end point. Indicates that all nodes have been visited; The computational geometry algorithm library CGAL is used to solve the generalized traveling salesman problem, and the final path obtained can achieve optimal coverage of the urban area.

[0031] Specifically, in S520, since the UAV can hover and turn in place, it is assumed that the trajectory of the UAV between the waypoints of the flight segment conforms to a straight line, and the speed ramp function is used to parameterize the planned path. Figure 6 and 7 As shown, the corresponding coverage results generated for a given map and the final generated path diagram.

[0032] The above-mentioned UAV coverage path planning method for urban areas decomposes the urban area into multiple feasible obstacle-free subspaces, proposes an optimization-based coverage path planning algorithm, and can achieve optimal coverage path planning in urban areas. It has the advantages of high flexibility, low computational complexity, strong robustness, and high safety.

[0033] The above is a detailed introduction to a UAV coverage path planning method for urban areas provided by the present invention. This article uses specific examples to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the core idea of ​​the present invention. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present invention, several improvements and modifications can be made to the present invention, and these improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A method for planning a UAV coverage path for an urban area, characterized in that: The method comprises the following steps: S100: Determine the starting point and end point coordinates of the UAV coverage mission, and build a map model based on the terrain database of the area where the flight mission is located; S200: Establishing a height model for the UAV coverage flight, preprocessing the grid map according to the height model, and obtaining a feasible space for the coverage task; S300: Design an integer programming method for set partitioning to decompose the target area into multiple feasible subspaces that do not contain obstacles, and obtain a spatial decomposition diagram of the task area; S400: Based on the space decomposition graph, construct an optimization problem with the cost of minimizing the path length and the starting point, end point and covering direction of the subspace coverage planning as output; S500: The optimization problem is equivalently converted into a generalized traveling salesman problem for solution, a path covering each subspace is generated, and path planning covering the entire target area is completed.

2. The method according to claim 1, characterized in that S100 includes: S110: Determine the starting point and end point coordinates of the UAV flight according to the UAV coverage mission requirements and initial conditions; S120: Generate a digital topographic map of the target area using the terrain data, clearly define the horizontal and vertical boundaries of the coverage task, and the vertex positions and height information of obstacles in the area; S130: Convert digitized topographic map into raster map: ; in Represents the target area, Represents the horizontal and vertical coordinates of the drone, Represents the height of the grid point, represents the entire set of real numbers.

3. The method according to claim 2, characterized in that S200 includes: S210: Establish UAV altitude model; S220: Before executing the mission, adjust the altitude of the drone's coverage flight according to the mission requirements and the target area environmental information, and pre-process the raster map of the area according to the flight altitude, wherein the obstacle area greater than the drone's flight altitude is regarded as an inaccessible area, and the obstacle area less than the drone's flight altitude is regarded as a reachable area, and generate a target area raster map for coverage planning.

4. The method according to claim 3, characterized in that S210 is specifically: ; in Represents the altitude of the drone, Represents the requirement for clarity, represents the clarity weight, Represents the wind factor in the target area, represents the safety weight, It is the standard flight altitude of drones; The specific processing process in S220 is as follows: ; in represents the flight altitude of the drone, and k represents the altitude of the drone. The accessibility of Indicates that the drone is unreachable. Represents drone reachable; The new grid map is as follows: ; According to the query The value of can determine the corresponding Obstacle situation at the location.

5. The method according to claim 4, characterized in that S300 includes: S310: Extending edges of obstacles with inner angles less than 180 degrees until they hit a map boundary or other obstacles, and extending the edges to divide the target space into multiple subspaces; S320: merging the divided subspaces by using an integer programming algorithm for set partitioning, thereby reducing the number of subspaces; S330: Decomposing the target area into a plurality of sub-areas that do not contain obstacles by solving an integer programming problem.

6. The method according to claim 5, characterized in that The integer programming model for set partitioning in S320 is specifically: ; ; ; ; in By strategy The width of the generated convex polygon, represents the selected strategy, Represents all strategies that can be merged, , Indicates whether to select a strategy , 1 if selected, 0 otherwise, represents the subspace, Represents a collection of subspaces. If the merging strategy Include Sub-Polygons ,but is 1 if the value is true, otherwise it is 0.

7. The method according to claim 6, characterized in that S400 includes: S410: Based on the obtained task area spatial decomposition diagram, it is necessary to determine the starting point, end point and covering direction of the subspace, wherein the direction selection depends on the direction of each edge of the subspace, that is: ,in Represents the covering direction, F represents all the edges of the subspace, and the starting and ending points are the vertices of each subspace, that is: ,in , Represent the starting point and the end point respectively, and V represents all the nodes in the subspace; S420: Establish an optimization problem with minimizing the path length as the cost and taking the starting point, end point and coverage direction of the subspace coverage planning as the output.

8. The method according to claim 7, characterized in that The optimization problem of the path cost in S420 is specifically: ; in, is the total number of sub-regions decomposed, ) is the The sub-regions cover the coordinates of the path end points, ) is the The sub-regions cover the coordinates of the path starting point, Represents the cost of transitioning between different subspaces.

9. The method according to claim 8, characterized in that S500 includes: S510: Define adjacency graph: ,in represents the set of subspaces, Represents the access order, that is, from arrive The order in which access is performed; each access sequence There is a arrive The shortest path , use the existing graph search algorithm to obtain the distance of the shortest path; S520: The path is planned using the speed ramp function parameterization. The total cost of the trajectory is the sum of the costs of all segmented trajectories. The cost of each segment of the trajectory is The value depends on the distance between the two points 、Maximum speed of drone and maximum acceleration : ; in , , is the time to accelerate to maximum speed, is the distance the drone travels when accelerating, is the maximum speed, is the maximum acceleration, For arrive The shortest path; since the starting and ending points of the area coverage do not coincide, the cost matrix is asymmetric, that is ; S530: Based on S300, the vertex set of each subspace is fixed. By selecting appropriate starting and ending points and the order of visiting the subspaces to ensure the minimization of the transition time, the following generalized traveling salesman problem is constructed: ; in Represents the total cost, The first visit represents the starting point. The last visited one is the end point. Indicates that all nodes have been visited; The computational geometry algorithm library CGAL is used to solve the generalized traveling salesman problem, and the final path obtained can achieve optimal coverage of the urban area.

Citation Information

Patent Citations

  • Unmanned aerial vehicle flight planning method and device, computer equipment and storage medium

    CN113268079A

  • Air base station deployment and path planning method under flight path constraint

    CN118092505A

  • Optimization algorithm for three-dimensional coverage trajectory of unmanned aerial vehicle in multi-terrain environment

    CN118706118A

  • Three-dimensional model-based coverage path planning method for unmanned aerial vehicles

    US20220215621A1

Cited By

  • Unmanned aerial vehicle path planning method and system based on reinforcement learning

    CN120403661A

  • A UAV path planning method and system based on reinforcement learning

    CN120403661B

  • Distributed urban and rural planning decision support method and system

    CN121660208A

  • Planar region coverage path generation method, computer equipment and storage medium

    CN121903099A