An adaptive path planning method for unmanned aerial vehicle regional full coverage

Abstract

The application discloses an adaptive path planning method for unmanned aerial vehicle (UAV) area full coverage, and relates to the technical field of UAV path planning. In order to solve the problems that the existing grid planning algorithm in the mountainous terrain or disaster environment with unknown map is insufficient in map coverage due to too large grid division, and is increased in operation amount and lengthened in path due to too small grid division, the application adopts an adaptive grid division method based on a perception radius, combines the grid length setting with the perception radius of the UAV, and optimizes the grid size through a dynamic adjustment function; for the edge rough obstacle, small gaps are processed through independent area division and setting of an obstacle inflation radius; when the path falls into a dead zone, a path escape algorithm is adopted to select a correct blank node as an escape target. The application is suitable for full coverage path planning of the UAV in a complex environment.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) path planning technology, and in particular to an adaptive path planning method for full regional coverage of UAVs. Background Technology

[0002] With the widespread application of drone technology in fields such as agricultural plant protection, regional inspection, and disaster monitoring, higher requirements are being placed on the autonomous coverage capabilities of drones.

[0003] Existing full-coverage path planning schemes, especially grid-based methods, face several key technical challenges: 1. When the grid is too large, map coverage is insufficient, resulting in missed coverage areas; when the grid is too small, computational load increases significantly, leading to longer flight paths and reduced operational efficiency. 2. Traditional algorithms struggle to achieve complete coverage when facing obstacles with rough and complex edges, easily creating coverage blind spots. 3. During path planning, UAVs are prone to getting stuck in dead zones (positions surrounded by obstacles or already covered grids). If existing obstacle avoidance algorithms frequently conflict with the original planning algorithms, it can lead to duplicate coverage or traversing impassable nodes. These technical bottlenecks severely restrict the efficiency of full-coverage path planning for UAVs in complex environments, resulting in a high rate of path duplication in practical applications.

[0004] The aforementioned technical challenges ultimately lead to a situation where, when dealing with complex terrains such as mountainous areas or disaster environments where maps are unknown, the grid planning results of UAVs often suffer from two problems: grids that are too large result in insufficient map coverage, while grids that are too small result in increased computational load and longer paths. Both of these issues ultimately lead to a high rate of path repetition in the planned paths.

[0005] Therefore, how to improve the full-coverage path planning method of UAVs in complex terrain and reduce the repetition rate of the planned paths has become a research topic. Summary of the Invention

[0006] The embodiments of the present invention provide an adaptive path planning method for full coverage of UAV areas, which can improve the full coverage path planning method of UAVs in complex terrain and reduce the repetition rate of the planned path.

[0007] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0008] An adaptive path planning method for full area coverage by unmanned aerial vehicles (UAVs) includes:

[0009] S1. Dynamically adjust the grid length according to the drone's sensing radius;

[0010] S2. Based on the adjusted grid length, perform full-coverage path planning on the constructed 3D map, and then execute S3-S4 to optimize the path;

[0011] S3. Identify dead zones on the planned path and execute an escape plan to assist the drone in escaping the dead zones;

[0012] S4. Refine the edge grid of obstacles on the planned path.

[0013] Specifically, in S1, the methods for dynamically adjusting the grid length include: r is the sensing radius of the drone, and L is the grid length corresponding to the sensing radius of the drone. ω is a radius adjustment coefficient, which is used for dynamically adjusting parameters.

[0014] Specifically, in S2, the method for performing full-coverage path planning includes: using a coverage model for full-coverage path planning, wherein the coverage model is... , Where F is the coverage rate and S is the total number of complete scanned gratings. This represents the number of non-repeating raster cells in a complete scan. This represents the reduction in coverage repetition rate, where P represents the average number of coverage iterations of the existing method. This indicates the average number of times the system is covered.

[0015] During the 3D map construction process, a 3D raster map M of the UAV's working environment is generated, where, , Indicates the coordinates of the point. The grid state matrix, The map dimensions are defined as follows: i, j, k represent the x-axis, y-axis, and z-axis coordinate values, respectively; m, n', and q represent the dimensions of the map along the x-axis, y-axis, and z-axis, respectively. The state information of the raster is stored in a matrix. In this process, the grid state matrix is ​​obtained. .

[0016] Further according to The drone's direction is selected on the constructed 3D map, whereby... This represents the set of 8 connected neighbors in the full coverage algorithm. .

[0017] Specifically, in S3, the escape scheme to assist the drone in escaping the dead zone includes: when the drone is detected to be surrounded by obstacles or covered grids, using the A* algorithm to find the nearest uncovered grid node and generate an escape path; wherein, during the process of generating the escape path using the A* algorithm, the heuristic cost from the current node n to the target t is estimated. A represents the working range of a single drone. , Weights for choosing the direction of the two nodes. This represents the x-coordinate of the current node. This represents the y-coordinate of the current node. This represents the x-axis coordinate of the target node. Represents the y-axis coordinate of the target node;

[0018] When generating an escape path fails, a new escape path is generated based on an improved heuristic function, where the improved heuristic function is: , This indicates the current location and status of the drone. This represents the actual path cost from the starting point s to node n: ;in The x-axis coordinates representing the starting point. The y-axis coordinate of the starting point. Let be the comprehensive cost function corresponding to the target node, where , For the current node, For the target node, This represents the Euclidean distance from the current node to the target node. This indicates the location parameters of obstacles near the target node. The cost of the target node's height. They are respectively the corresponding , and The weighting coefficients.

[0019] Specifically, S4 includes: adjusting the grid length using a multi-level threshold judgment mechanism, including: ignoring gaps whose gap size is less than or equal to the first-level threshold; for gaps that cannot be ignored, if the second-level threshold ≤ gap size ≤ original grid length, then the surrounding grid of the obstacle is initially divided; for gaps that have been initially divided, if the third-level threshold ≤ gap size < second-level threshold, then the edge areas that still cannot be covered are further divided; wherein, the multi-level threshold judgment mechanism sets multi-level gap size thresholds for adjusting the sensing radius; the first-level threshold is the safe radius of the UAV, the second-level threshold is 1 / 2 of the original grid length, and the third-level threshold is 1 / 4 of the original grid length.

[0020] Further subdivision of edge areas that still cannot be covered includes: for gaps smaller than the third-level threshold, the gaps are incorporated into the obstacle area by increasing the obstacle expansion radius; for gaps larger than the third-level threshold, they are divided into independent grids.

[0021] The adaptive path planning method for full coverage of UAV areas provided in this invention adopts adaptive grid division based on perception radius, combining grid length setting with the UAV's perception radius, and optimizing grid size through a dynamic adjustment function. For obstacles with rough edges, small gaps are handled by independent region division and setting obstacle expansion radius. When the path gets stuck in a dead zone, an escape algorithm is used to select the correct blank node as the escape target. In particular, it addresses the problems of existing grid planning algorithms in mountainous terrain or disaster environments with unknown maps, where excessively large grid divisions lead to insufficient map coverage, while excessively small grid divisions lead to increased computation and longer paths. This method improves the full coverage path planning method for UAVs in complex terrain and reduces the repetition rate of the planned paths. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A technical roadmap provided for embodiments of the present invention;

[0024] Figure 2 A flowchart of full-coverage path planning provided in an embodiment of the present invention;

[0025] Figure 3 A schematic diagram of the program flow for the adaptive grid division method provided in an embodiment of the present invention;

[0026] Figure 4 This is a diagram illustrating the obstacle edge segmentation effect of the algorithm provided in an embodiment of the present invention.

[0027] Figure 5 A flowchart illustrating the dead zone escape method provided in an embodiment of the present invention;

[0028] Figure 6 This is a diagram illustrating the algorithm's dead zone escape effect provided in an embodiment of the present invention.

[0029] Figure 7 , Figure 8 This is a diagram showing the planning results of the full coverage algorithm provided in an embodiment of the present invention;

[0030] Figure 9 This is a path coverage diagram for existing technology cases;

[0031] Figure 10 This is a schematic diagram of path coverage provided in an embodiment of the present invention. Detailed Implementation

[0032] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Embodiments of the present invention will be described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” “the,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in the specification of the present invention means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or couplings. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the meaning consistent with their meaning in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0033] This invention provides an adaptive path planning method for full coverage of unmanned aerial vehicles (UAVs). The main design idea is to adjust the grid length by changing the sensing radius to reduce redundant scanning; simultaneously, for irregularly shaped obstacles, the surrounding grid planning scheme is modified to better achieve edge coverage. The method includes:

[0034] S1. Dynamically adjust the grid length according to the drone's perception radius; wherein, dynamically adjust the grid length according to the drone's perception radius; the grid is divided by a planning function so that the most suitable grid size can be planned according to different drone recognition ranges;

[0035] S2. Based on the adjusted grid length, perform full-coverage path planning on the constructed 3D map, followed by path optimization in steps S3-S4. Specifically, full-coverage path planning is performed on the entire 3D map, and a full-coverage planning algorithm is run based on map information, ultimately yielding a path planning algorithm with a coverage rate exceeding 95%. For example... Figure 1 ,2 As shown, the overall process of this implementation is divided into a partitioning stage, a coverage stage, and an escape stage. First, the map is divided into grids based on the map size and the drone's perception radius. Then, a full coverage algorithm is run. If the drone gets stuck in a dead zone, an escape algorithm is run to help it escape. Finally, a complete full coverage route is obtained.

[0036] S3. Identify dead zones on the planned path and generate solutions; dead zone escape refers to the situation where the UAV is surrounded by obstacles and has already passed the path when executing the full coverage algorithm. It is one of the problems that arise during path planning. The best time to solve dead zone escape is before the UAV actually takes off. In practical applications, it will be solved during the path planning process.

[0037] S4. Refine the edge grid of obstacles on the planned path to ensure coverage and flight logic.

[0038] In this embodiment, the method for dynamically adjusting the grid length in S1 includes: 'r' represents the drone's sensing radius, which can be understood as a value determined by the drone in actual operation based on its scanning capabilities and map requirements. It is an inherent data point of the drone and can be set according to actual needs. 'L' represents the grid length corresponding to the drone's sensing radius. The grid length can be dynamically adjusted according to the drone's sensing radius. In the optimal solution, the grid length is set to half of the sensing radius, and the grid is divided using a planning function. This ensures that the range of a single grid cell represents the actual scanning range of the drone, making the output results more intuitive. For dynamic adjustment parameters, their purpose is to assign values ​​according to the set values ​​of each threshold and participate in the determination of multi-level threshold grids. The value of the dynamic adjustment parameter can be set according to the positional relationship of obstacles. ω is the radius adjustment coefficient used to manually correct the grid division on the existing basis (its value is generally set to 1).

[0039] Furthermore, the main optimization strategy for dynamically adjusting the grid length is as follows: when encountering complex edge obstacles, excessively small gaps are designated as obstacle regions by expanding the radius, while larger gaps are divided into separate grids and their surrounding paths are optimized. Specifically, the dynamic adjustment function... Taking into account both environmental complexity and obstacle distribution, the grid size is appropriately increased in simple environments and decreased in complex environments to optimize coverage efficiency; ω is the radius adjustment coefficient, which can be changed... The value is used to fine-tune the raster. The process of dynamically adjusting the raster length includes:

[0040] A multi-level threshold judgment mechanism is used to adjust the grid length, for example... Figure 3-4As shown, multiple gap size thresholds are set. The first threshold is the drone's safe radius, used to determine whether to ignore small gaps. If the gap size is less than or equal to the first threshold, it is considered a small gap and is ignored. The second threshold is half the grid length. If the second threshold is less than or equal to the gap size and the original grid length, the grid around the obstacle is initially divided. The third threshold is one-quarter of the grid length, used to further divide the edge areas that still cannot be covered. If the third threshold is less than or equal to the gap size and less than the second threshold, further division is performed.

[0041] The invention employs a layered processing approach for gaps around obstacles. For gaps smaller than the third-level threshold, the obstacle's expansion radius is increased to incorporate the gap into the obstacle area. For gaps larger than or equal to the third-level threshold, independent grids are created, and surrounding paths are replanned to ensure complete coverage. The edge obstacle processing module is used to independently divide obstacles with irregular edges. By setting the obstacle expansion radius, it ignores the interference of small gaps on path planning, while independently setting grids for larger gaps and adjusting the layout of adjacent grids. This invention's processing solution for rough-edged obstacles, through obstacle expansion radius setting and independent grid division, effectively handles complex edge situations, increasing edge area coverage from 85%-90% in traditional methods to over 98%, significantly reducing coverage blind spots.

[0042] This processing method, which utilizes multi-level gap size thresholds, avoids path fragmentation caused by too many small grids while ensuring complete coverage of important gap areas. In practical applications, coverage is improved by approximately 8%-12%, and path repetition rate is reduced by approximately 5%-8%. Figure 7 , Figure 8 The diagram shows the planning results of the full coverage algorithm, where... Figure 7 This is a diagram showing the planning results of the full coverage algorithm in a current embodiment. Figure 8 The full coverage algorithm planning result diagram provided in the embodiments of the present invention can be obtained through... Figure 7 and Figure 8 The comparison shows that the method used in this embodiment effectively reduces the full-coverage path length when facing complex terrain. For complex edge obstacles, a layered processing strategy is adopted: for small gaps smaller than a threshold (usually set to 1 / 4 of the UAV's perception radius), the obstacle expansion radius is set to include them in the obstacle area; for larger gaps, independent grids are divided and the surrounding paths are replanned to ensure full coverage without omissions. When the UAV is detected to be trapped in a dead zone, a search algorithm is initiated, using the nearest uncovered grid as the target node, and comprehensively considering path length, turning cost, and coverage efficiency to generate the optimal escape path.

[0043] In this embodiment, the method for performing full-coverage path planning in S2 includes:

[0044] Using a coverage model for full-coverage path planning, compared to the fixed-grid partitioning method, the coverage model, while ensuring full coverage, provides... , Where F is the coverage rate and S is the total number of complete scanned gratings. This represents the number of non-repeating raster cells in a complete scan. This represents the reduction in coverage repetition rate, where P represents the average number of coverage iterations of the existing method. This indicates the average number of times the drone covers the area. For example, using nodes that the drone has passed through or may pass through (e.g., during simulated flight on a planned path) as the center, and the drone's sensing radius as the scanning radius, all coordinates within this range are recorded as scanned points. Figure 4 As shown, considering the unavoidable duplicate coverage during full coverage, this patent first calculates the theoretically optimal average coverage value of 2.2 times per grid for a single UAV range (a range of 20m×20m). Existing algorithms (not considering the perception radius) have an average coverage of 7.7 times per grid on maps containing obstacles. The method used in this patent has an average coverage of 5.4 times per grid on the same map containing obstacles. In comparison, the method used in this patent reduces the coverage duplication rate by about 29.9% while ensuring full coverage, effectively solving the contradiction between coverage and computational load in traditional methods. Figure 9 This is a path coverage diagram for existing technology cases. Figure 10 The path coverage diagram provided in the embodiment of the present invention shows that, by comparison, the embodiment of the present invention effectively reduces the repeated scanning of paths.

[0045] Furthermore, during the 3D map construction process, a 3D raster map M of the UAV's working environment is generated, wherein... , Indicates the coordinates of the point. The grid state matrix, For map dimensions, i, j, k represent x-axis, y-axis, and z-axis coordinate values ​​respectively, and m, n', and q represent the map dimensions along the x-axis, y-axis, and z-axis respectively.

[0046] Store the state information of each grid cell into a matrix. In the matrix, 0 represents uncovered, 1 represents covered, and inf represents obstacles, resulting in the grid state matrix. .

[0047] according to The drone's direction is selected on the constructed 3D map, whereby... The set of 8 connected neighbors in the full coverage algorithm is represented as: .

[0048] In this embodiment, as Figure 5 , 6 As shown, when a drone gets stuck in a dead zone, the A* algorithm, combined with a CNN neural network, selects an uncovered blank node as the escape target, generating the optimal escape path and avoiding path conflicts and duplicate coverage. S3 includes: when the drone is detected to be surrounded by obstacles or covered grids, the A* algorithm is used to find the nearest uncovered grid node and generate an escape path; wherein, during the process of generating the escape path using the A* algorithm, the heuristic cost from the current node n to the target t is estimated. A represents the working range of a single drone. , Weights for choosing the direction of the two nodes. This represents the x-coordinate of the current node. This represents the y-coordinate of the current node. This represents the x-axis coordinate of the target node. This represents the y-axis coordinate of the target node.

[0049] Since the A* algorithm alone cannot completely solve the problem of getting stuck, an improvement is made to the existing A* algorithm. This improvement addresses the issue of the A* algorithm's inability to escape in situations with complex local terrain. Let the current node position be... The candidate target node position is The comprehensive cost function corresponding to the target node: ,in: This represents the Euclidean distance from the current node to the target node. Indicates the location parameters of obstacles near the target node; The cost of the target node's height; Here are the weighting coefficients. Therefore, the improved heuristic function can be obtained as follows: ,in: This indicates the current location and status of the drone. This indicates the next node position status of the drone. and The two modified cost functions are based on the paths planned by the above-mentioned optimized A* algorithm, enabling them to complete the planning in a shorter route and in a faster time.

[0050] In this embodiment, S4 includes: adjusting the grid length using a multi-level threshold judgment mechanism, wherein multi-level gap size thresholds are set; the first level threshold is the safe radius of the UAV, and gaps with gap sizes less than or equal to the first level threshold are ignored; for gaps that cannot be ignored, if the gap size is less than the second level threshold, the grid around the obstacle is initially divided, and the second level threshold is 1 / 2 of the grid length; for gaps that have been initially divided, if the gap size is less than or equal to the third level threshold, the edge areas that still cannot be covered are further divided, and the second level threshold is 1 / 4 of the grid length.

[0051] The further division of the edge areas that still cannot be covered includes: for gaps with a size smaller than the third-level threshold, the gaps are incorporated into the obstacle area by increasing the expansion radius of the obstacle; for gaps with a size larger than the third-level threshold, they are divided into grids independently.

[0052] In practical applications, a hierarchical decision-making architecture can be adopted. The bottom layer is the grid partitioning layer, responsible for environment modeling; the middle layer is the path planning layer, generating coverage paths; and the top layer is the decision control layer, handling abnormal situations and mode switching. This hierarchical architecture gives the algorithm good scalability and adaptability, enabling it to cope with various complex operating environments. In practical applications, the algorithm can achieve a full coverage task completion rate of over 96% in complex environments, an improvement of 10%-15% compared to traditional methods. In this embodiment, an environmental map is dynamically built based on environmental characteristics and UAV capabilities; then, special processing is applied to complex obstacle areas to ensure edge coverage integrity; during path execution, the UAV status is monitored in real time, and an escape path is quickly generated when stuck in a dead zone. The entire process forms a closed-loop control of "planning-execution-monitoring-adjustment," ensuring the efficient and reliable completion of the full coverage task. This solves various problems existing in traditional full coverage path planning algorithms in complex environments, significantly improving the autonomous operation capability and efficiency of UAVs. It ensures the integrity of full coverage while significantly improving planning efficiency and path quality, making it particularly suitable for autonomous UAV operation tasks in complex environments, with broad applicability and significant practical value.

[0053] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on its differences from other embodiments. In particular, the device embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. The above descriptions are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An adaptive path planning method for full area coverage by unmanned aerial vehicles (UAVs), characterized in that, include: S1. Dynamically adjust the grid length according to the drone's sensing radius; S2. Based on the adjusted grid length, perform full-coverage path planning on the constructed 3D map, and then execute S3-S4 to optimize the path; S3. Identify dead zones on the planned path and execute an escape plan to assist the drone in escaping the dead zones; S4. Refine the edge grid of obstacles on the planned path; In S2, the methods for performing full-coverage path planning include: using a coverage model for full-coverage path planning, wherein the coverage model is... , Where F is the coverage rate and S is the total number of complete scanned gratings. The number of non-repeating raster cells in the complete scan. This represents the reduction in coverage repetition rate, where P represents the average number of coverage iterations of the existing method. Indicates the average number of times coverage is achieved; In S3, the escape schemes used to assist drones in escaping dead zones include: When the drone is detected to be surrounded by obstacles or covered by grids, the A* algorithm is used to find the nearest uncovered grid node and generate an escape path; where the heuristic cost from the current node n to the target t is calculated. A represents the working range of a single drone. , Weights for choosing the direction of the two nodes. This represents the x-coordinate of the current node. This represents the y-coordinate of the current node. This represents the x-axis coordinate of the target node. Represents the y-axis coordinate of the target node; When generating an escape path fails, a new escape path is generated based on an improved heuristic function, where the improved heuristic function is: n is the current node. This represents the actual path cost from the starting point s to the current node n: ;in The x-axis coordinates of the starting point. The y-coordinate of the starting point. This is the comprehensive cost function for the corresponding target node; , For the current node, For the target node, This represents the Euclidean distance from the current node to the target node. This indicates the location parameters of obstacles near the target node. The cost of the target node's height. They are respectively the corresponding , and The weighting coefficients.

2. The method according to claim 1, characterized in that, In S1, the methods for dynamically adjusting the grid length include: r is the sensing radius of the drone, and L is the grid length corresponding to the sensing radius of the drone. ω is a radius adjustment coefficient, which is used for dynamically adjusting parameters.

3. The method according to claim 1, characterized in that, S2 includes: During the 3D map construction process, a 3D raster map M of the UAV's working environment is generated, where, , Indicates the coordinates of the point. The grid state matrix, For map dimensions, i, j, k represent x-axis, y-axis, and z-axis coordinate values ​​respectively, and m, n', and q represent the map dimensions along the x-axis, y-axis, and z-axis respectively. Store the state information of the raster into a matrix. In this process, the grid state matrix is ​​obtained. .

4. The method according to claim 3, characterized in that, Also includes: according to The drone's direction is selected on the constructed 3D map, whereby... This represents the set of 8 connected neighbors in the full coverage algorithm. .

5. The method according to claim 2, characterized in that, S4 includes: adjusting the grid length using a multi-level threshold judgment mechanism, including: Ignore gaps whose size is less than or equal to the first-level threshold; For gaps that cannot be ignored, if the second-level threshold ≤ gap size ≤ original grid length, then the surrounding grid of the obstacle is initially divided. For the gaps that have been initially divided, if the third-level threshold ≤ gap size < second-level threshold, then the edge areas that still cannot be covered are further divided. The multi-level threshold judgment mechanism includes a multi-level gap size threshold for adjusting the sensing radius; the first level threshold is the safe radius of the UAV, the second level threshold is 1 / 2 of the original grid length, and the third level threshold is 1 / 4 of the original grid length.

6. The method according to claim 5, characterized in that, Further subdivision of edge areas that still cannot be covered includes: for gaps with a gap size smaller than the third-level threshold, the gaps are incorporated into the obstacle area by increasing the expansion radius of the obstacle; For gaps larger than the third-level threshold, separate grids are created.

Images