Path planning method and system based on skeleton nodes and probabilistic road diagram

Abstract

The invention relates to a path planning method and system based on skeleton nodes and a probability road map. The method comprises the following steps: acquiring an obstacle map comprising path obstacles; performing expansion operation on the edge of the obstacle map and the path obstacle to obtain a map skeleton; extracting line intersection points in the map skeleton to obtain skeleton nodes of the map skeleton; constructing a collision-probability-free road map for avoiding obstacles based on the skeleton nodes; and searching the non-collision probability road map based on a shortest path priority principle to obtain a shortest target path from the path starting point to the path end point, thereby solving the problems of low planning efficiency and high efficiency caused by easy generation of redundant path nodes or invalid search in a complex application scene in the existing traditional algorithm. And it is difficult to quickly generate an optimal path meeting actual requirements.

Description

Technical Field

[0001] This application relates to the field of navigation technology, and in particular to a path planning method and system based on skeleton nodes and probabilistic road maps. Background Technology

[0002] Path planning is one of the core technologies in fields such as robotics, autonomous driving, drones, and logistics scheduling. Its essence is to find an "optimal path" (such as the shortest distance, shortest time, and lowest energy consumption) for a mobile subject (robot, vehicle, etc.) from the starting point to the target point under specific constraints (such as environmental obstacles, movement restrictions, energy consumption / time costs).

[0003] However, when faced with complex application scenarios (such as large map scale, numerous trivial obstacles in the environment, narrow passages or narrow work areas, or large areas of barrier-free space in the environment), traditional algorithms are prone to generating redundant path nodes or invalid searches, resulting in low planning efficiency and difficulty in quickly generating the optimal path that meets actual needs. Summary of the Invention

[0004] This application provides a path planning method and system based on skeleton nodes and probabilistic road maps, which can solve the problem that existing traditional algorithms are prone to generating redundant path nodes or invalid searches when facing complex application scenarios, resulting in low planning efficiency and difficulty in quickly generating the optimal path that meets actual needs.

[0005] Firstly, this application provides a path planning method based on skeleton nodes and probabilistic road maps, including: Obtain an obstacle map that includes path obstacles; The edges of the obstacle map and the path obstacles are expanded to obtain the map skeleton; Extract the line intersections from the map skeleton to obtain the skeleton nodes of the map skeleton; A collision-free probability roadmap for obstacle avoidance is constructed based on the skeleton nodes; Based on the shortest path priority principle, the collision-free probability roadmap is searched to obtain the shortest target path from the path start point to the path end point.

[0006] In some embodiments, the step of expanding the edges of the obstacle map and the edges of the obstacles to obtain a map skeleton includes: Convert the obstacle map into an obstacle grid map; Define the edges of the obstacle grid map as additional obstacles, and perform an expansion operation on the additional obstacles and the path obstacles; Detect the expanded distance between the additional obstacle and the path obstacle after the expansion operation is performed; If the distance after dilation is equal to the grid number threshold, then the dilation operation is stopped and the map skeleton is obtained.

[0007] In some embodiments, converting the obstacle map into an obstacle grid map includes: Set the conversion resolution for converting the obstacle map into a raster map; The obstacle map is converted into the obstacle grid map based on the conversion resolution.

[0008] In some embodiments, extracting line intersections from the map skeleton to obtain the skeleton nodes of the map skeleton includes: The map skeleton is simplified into a line skeleton map; Extract the line intersections from the line skeleton map to obtain the skeleton nodes of the map skeleton.

[0009] In some embodiments, constructing a collision-free probability roadmap for obstacle avoidance based on the skeleton nodes includes: Connect any two of the skeleton nodes to obtain a global path graph that includes all paths between nodes; The global path map is filtered based on the path obstacles to obtain the collision-free probability road map.

[0010] In some embodiments, filtering the global path map based on the path obstacles to obtain the collision-free probability road map includes: Import the path obstacles into the global path graph; Determine whether the path between the nodes crosses any obstacles along the path; If the path between nodes crosses the path obstacle, then the path between nodes is deleted to obtain the collision-free probability roadmap.

[0011] In some embodiments, before connecting any two skeleton nodes to obtain a global path graph including all paths between nodes, the method further includes: Obtain the starting point and the ending point of the path; The starting point and the ending point of the path are defined as skeleton nodes.

[0012] In some embodiments, the step of searching the collision-free probability roadmap based on the shortest path priority principle to obtain the shortest target path from the path start point to the path end point includes: The graph search algorithm is used to search for multiple collision-free feasible paths from the starting point of the path to the ending point of the path, and the collision-free path length of the multiple collision-free feasible paths is calculated. Based on the shortest path priority principle, multiple collision-free path lengths are selected to obtain the shortest path length; Search for the collision-free feasible path corresponding to the shortest path length to obtain the shortest target path.

[0013] Secondly, embodiments of this application provide a path planning system based on skeleton nodes and probabilistic road maps, applied to the path planning method based on skeleton nodes and probabilistic road maps as described in any one of the first aspects, comprising: The map acquisition module is used to acquire an obstacle map, including path obstacles; The expansion module is used to expand the edges of the obstacle map and the path obstacles to obtain the map skeleton. The node extraction module is used to extract the line intersections in the map skeleton to obtain the skeleton nodes of the map skeleton. The road map construction module is used to construct a collision-free probability road map for obstacle avoidance based on the skeleton nodes; The path filtering module is used to search the collision-free probability roadmap based on the shortest path priority principle to obtain the shortest target path from the path start point to the path end point.

[0014] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the path planning method based on skeleton nodes and probabilistic road maps as described in any one of the first aspects.

[0015] The technical solutions provided in this application have the following advantages compared with the prior art: The path planning method and system based on skeleton nodes and probabilistic road maps provided in this application embodiment obtains an obstacle map including path obstacles; expands the edges of the obstacle map and the path obstacles to obtain a map skeleton; extracts the line intersections in the map skeleton to obtain the skeleton nodes of the map skeleton; constructs a collision-free probabilistic road map for obstacle avoidance based on the skeleton nodes; searches the collision-free probabilistic road map based on the shortest path priority principle to obtain the shortest target path from the path start point to the path end point. By constructing a collision-free probabilistic road map based on the expanded map skeleton, the planned path is far away from obstacles, ensuring the collision-free attribute of the generated path from the underlying logic. By extracting the map skeleton and skeleton nodes, the complex two-dimensional map path search problem is transformed into a topological graph search problem based on skeleton nodes, which greatly compresses the path search space. It can solve the problem that existing traditional algorithms are prone to generating redundant path nodes or invalid searches when facing complex application scenarios, resulting in low planning efficiency and difficulty in quickly generating the optimal path that meets actual needs. Attached Figure Description

[0016] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] One or more embodiments are illustrated by way of example with reference numerals in the accompanying drawings. These illustrations do not constitute a limitation on the embodiments. Elements with the same reference numerals in the drawings are denoted as similar elements. Unless otherwise stated, the figures in the drawings are not to be limited by scale.

[0019] Figure 1 A flowchart of a path planning method based on skeleton nodes and probabilistic road maps provided in an embodiment of this application; Figure 2 A schematic diagram a of skeleton extraction provided for an embodiment of this application (including an obstacle map of path obstacles); Figure 3 Schematic diagram b of skeleton extraction (first dilation operation) provided for an embodiment of this application; Figure 4 Schematic diagram c (second dilation operation) of skeleton extraction provided in an embodiment of this application; Figure 5 A schematic diagram d (map skeleton and skeleton nodes) for skeleton extraction provided in an embodiment of this application. Figure 6 The location provided for one embodiment of this application Figure 1 skeleton; Figure 7 The location provided for one embodiment of this application Figure 2 skeleton; Figure 8 The location provided for one embodiment of this application Figure 3 skeleton; Figure 9 The location provided for one embodiment of this application Figure 1 Skeleton nodes; Figure 10 The location provided for one embodiment of this application Figure 2 Skeleton nodes; Figure 11 The location provided for one embodiment of this application Figure 3 Skeleton nodes; Figure 12 The location provided for one embodiment of this application Figure 1 Probability road map; Figure 13 The location provided for one embodiment of this application Figure 2 Probability road map; Figure 14 The location provided for one embodiment of this application Figure 3 Probability road map; Figure 15 This is a collision-free probability roadmap provided in one embodiment of this application; Figure 16 The location provided for one embodiment of this application Figure 1 Path planning; Figure 17 The location provided for one embodiment of this application Figure 2 Path planning; Figure 18 The location provided for one embodiment of this application Figure 3 Path planning; Figure 19 A schematic diagram of Dijkstra's algorithm provided in an embodiment of this application; Figure 20 This is a schematic diagram of a path planning system based on skeleton nodes and probabilistic road maps provided in an embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] The following disclosure provides numerous different embodiments or examples for implementing various structures of the invention. To simplify the disclosure, specific examples of components and arrangements are described below. These are merely examples and are not intended to limit the scope of the invention. Furthermore, reference numerals and / or letters may be repeated in different examples. Such repetition is for simplification and clarity and does not in itself indicate a relationship between the various embodiments and / or arrangements discussed.

[0022] Path planning is one of the core technologies in fields such as robotics, autonomous driving, drones, and logistics scheduling. Its essence is to find an "optimal path" (such as the shortest distance, shortest time, and lowest energy consumption) for a mobile subject (robot, vehicle, etc.) from the starting point to the target point under specific constraints (such as environmental obstacles, movement restrictions, energy consumption / time costs).

[0023] The core objective of path planning is to plan an effective movement path from the starting point to the target point for a drone, aircraft, mobile robot, or other subject in a known environment map. This path must meet the basic constraint of "obstacle avoidance" and be highly efficient, ensuring that the device can quickly respond to task requirements and complete the operation smoothly. The speed and reliability of path planning directly determine the operational efficiency and practical value of mobile devices. If path planning takes too long, the device will be stuck in a waiting state, which will not only delay the task progress but may also lead to problems such as operation interruption and energy waste.

[0024] However, in complex real-world applications, existing path planning methods still face significant efficiency bottlenecks, struggling to balance speed and obstacle avoidance effectiveness. For example: 1) Large map scales (e.g., city-level delivery, large-area inspection) lead to an exponential increase in the path search space, resulting in a surge in planning computation; 2) Numerous minor obstacles in the environment (e.g., scattered tools in industrial workshops, clutter in homes) necessitate frequent adjustments to path nodes to avoid obstacles, increasing planning complexity; 3) Narrow passages and confined work areas (e.g., mine tunnels, warehouses with dense shelving) limit the feasible path domain, requiring meticulous searching for the optimal route; 4) Large, barrier-free areas in the environment (e.g., open squares, outdoor spaces) can lead to redundant path nodes or ineffective searches using traditional algorithms, resulting in low planning efficiency. These issues directly hinder existing path planning methods from quickly generating optimal paths that meet actual needs, severely restricting the application effectiveness and promotion potential of various mobile devices in complex scenarios.

[0025] Firstly, such as Figure 1-19 As shown, to address the aforementioned technical problems, this application provides a path planning method based on skeleton nodes and probabilistic road maps, including: S101: Obtain an obstacle map including path obstacles; S102: Perform an expansion operation on the edges of the obstacle map and the path obstacles to obtain the map skeleton; S103: Extract the line intersections in the map skeleton to obtain the skeleton nodes of the map skeleton; S104: Construct a collision-free probability roadmap for obstacle avoidance based on the skeleton nodes; S105: Search the collision-free probability roadmap based on the shortest path priority principle to obtain the shortest target path from the path start point to the path end point.

[0026] It should be noted that this method extracts the map skeleton by performing an expansion operation on the edges of the obstacle map and the obstacles on the path, and uses the skeleton intersections as core skeleton nodes. It does not require global sampling of the entire environment space. Compared with the random dense sampling of the environment by the traditional probabilistic road map (PRM) algorithm, the skeleton nodes directly anchor the key passage feature points of the environment, which can compress the path search space from the global space to the topological space composed of skeleton nodes, significantly reducing the computational load of path search. It is especially suitable for large-scale map scenarios such as city-level delivery and large-area area inspection, and can effectively meet the needs of real-time planning.

[0027] It should be noted that, as Figure 2 , 3 As shown in Figures 4 and 5, the expansion operation can maintain a safe distance between the map skeleton and the edge of the obstacle. The collision-free probabilistic roadmap constructed based on the skeleton nodes naturally has collision-free attributes, and no additional collision detection and path adjustment are required. For small obstacles such as scattered tools in industrial workshops and clutter in home environments, as well as narrow passages such as mine tunnels and warehouses with dense shelves, the skeleton nodes can accurately capture the feasible domain boundary of the environment, avoiding the problem of increased planning complexity caused by frequent adjustment of path nodes in traditional algorithms, and ensuring the feasibility of the path in narrow spaces.

[0028] It should be noted that in large, barrier-free areas such as open squares and outdoor venues, traditional algorithms are prone to generating redundant path nodes due to random sampling, resulting in tortuous paths and low planning efficiency. This method, based on the topology structure constructed by the skeleton nodes, can directly connect key access nodes within the barrier-free area. The generated paths are guided by the shortest priority principle, which avoids invalid search iterations and ensures the optimality and simplicity of the paths, thereby improving the operation efficiency of mobile devices.

[0029] In some embodiments, the step of expanding the edges of the obstacle map and the edges of the obstacles to obtain a map skeleton includes: Convert the obstacle map into an obstacle grid map; Define the edges of the obstacle grid map as additional obstacles, and perform an expansion operation on the additional obstacles and the path obstacles; Detect the expanded distance between the additional obstacle and the path obstacle after the expansion operation is performed; If the distance after dilation is equal to the grid number threshold, then the dilation operation is stopped and the map skeleton is obtained.

[0030] It should be noted that, as Figure 2 , 3As shown in Figures 4 and 5, by converting the obstacle map into the obstacle grid map, the continuous environmental space can be discretized into standardized grid units, enabling accurate quantitative description of obstacle edges and fragmented obstacle contours. Compared with the traditional vector map representation method, gridding can more intuitively define the boundary between obstacles and feasible regions, avoiding deviations in subsequent skeleton extraction caused by edge blurring, and providing a unified and computable basic data carrier for the expansion operation.

[0031] It should be noted that by defining the edges of the obstacle grid map as additional obstacles and performing an expansion operation together with path obstacles, a safe distance between the path and the environment boundary is ensured, as well as the collision-free nature of the path and internal obstacles. During the expansion process, the degree of expansion can be precisely controlled by detecting the distance between the expanded additional obstacles and path obstacles and using a grid number threshold (e.g., 1 grid) as the termination condition. This avoids problems such as over-expansion leading to loss of feasible domains or under-expansion leading to insufficient safety margins. The resulting map skeleton can accurately anchor key passageways in complex environments. After obtaining the skeleton nodes, a skeleton node set can be constructed based on these skeleton nodes. This facilitates the unified storage and management of skeleton nodes.

[0032] It should be noted that stopping the expansion operation usually means stopping the expansion in the direction where the distance after expansion is equal to the threshold of the number of grid cells (i.e., stopping the expansion in a directional manner). This allows for differentiated control based on the different obstacle distributions in different directions. For example, in a narrow passage scenario, the expansion distance of obstacles on both sides of the passage is likely to reach the threshold first. In this case, stopping the expansion in both directions in a directional manner can avoid the problem of the feasible region of the passage being over-compressed or even completely closed due to global expansion. In the direction of the barrier-free area, the expansion operation can continue to a reasonable range to ensure that the skeleton can completely cover the key passage paths of the environment, so that the final generated map skeleton is highly consistent with the topology of the real environment.

[0033] In some embodiments, converting the obstacle map into an obstacle grid map includes: Set the conversion resolution for converting the obstacle map into a raster map; The obstacle map is converted into the obstacle grid map based on the conversion resolution.

[0034] It should be noted that by pre-setting the conversion resolution, the size of the raster cells can be flexibly adjusted according to the needs of different application scenarios. For example, in narrow passage scenarios such as mine tunnels and warehouses with dense shelving, high-resolution raster cells can be selected to achieve fine representation of obstacle edges and narrow feasible regions, avoiding the loss of feasible regions due to excessively large raster granularity. In large-scale scenarios such as city-level delivery and large-scale inspection, low-resolution raster cells can be selected to significantly reduce the amount of raster data while ensuring that key obstacle features are not lost, thus balancing representation accuracy and computational efficiency. Furthermore, rasterization discretizes the continuous obstacle map into regular raster cells, transforming the positional information of obstacle edges and map boundaries into directly calculable raster coordinate data. Compared to traditional vector map representation methods, the rasterized map can complete the distance measurement and collision detection of the expansion operation without complex geometric calculations, significantly reducing the computational difficulty of subsequent additional obstacle definition and expansion distance threshold judgment, and improving the execution efficiency of the preprocessing stage.

[0035] In some embodiments, extracting line intersections from the map skeleton to obtain the skeleton nodes of the map skeleton includes: The map skeleton is simplified into a line skeleton map; Extract the line intersections from the line skeleton map to obtain the skeleton nodes of the map skeleton.

[0036] It should be noted that, as Figure 6 , 7 As shown in Figures 8, 9, 10, and 11, simplifying the map skeleton into a line skeleton map can eliminate redundant contour information generated after the expansion operation (such as redundant grid areas formed by the expansion of obstacle edges), retaining only the core passage path lines of the feasible environmental domain. This can transform the complex planar skeleton structure into a simple linear topology structure, significantly reducing the amount of target data extracted by subsequent nodes, avoiding node mis-extraction due to redundant information, and reducing the storage and computational overhead of skeleton nodes.

[0037] It should be noted that by using the intersection points of the lines in the skeleton map as skeleton nodes, key passage decision locations in complex environments can be accurately captured, such as entrances and exits of narrow passages, forks in the road where multiple paths intersect, and detour turning points surrounded by obstacles. These nodes naturally correspond to the core turning and selection logic in path planning. Compared with the random sampling nodes of traditional probabilistic road maps, the skeleton nodes extracted by this method are more targeted and can directly anchor the potential direction of the optimal path, providing efficient topological support for subsequent shortest path search.

[0038] In some embodiments, constructing a collision-free probability roadmap for obstacle avoidance based on the skeleton nodes includes: Connect any two of the skeleton nodes to obtain a global path graph that includes all paths between nodes; The global path map is filtered based on the path obstacles to obtain the collision-free probability road map.

[0039] It should be noted that, as Figure 12 , 13 As shown in Figures 14 and 15, by using skeleton nodes as vertices in the roadmap and directly connecting any two skeleton nodes to generate a global path graph, a global random sampling of the environment is not required. Compared to the random sampling-collision detection loop of the traditional PRM algorithm, the number of nodes in this step is determined by the environmental topological features rather than the sampling density, which can significantly reduce the size of the roadmap vertices and the computational cost of roadmap construction. Especially in large-scale city-level maps and large-area barrier-free areas, it can effectively avoid roadmap redundancy caused by redundant sampling points and improve the real-time performance of roadmap construction.

[0040] In some embodiments, filtering the global path map based on the path obstacles to obtain the collision-free probability road map includes: Import the path obstacles into the global path graph; Determine whether the path between the nodes crosses any obstacles along the path; If the path between nodes crosses the path obstacle, then the path between nodes is deleted to obtain the collision-free probability roadmap.

[0041] It should be noted that, as Figure 15 As shown, the path obstacles are directly imported into the global path graph. Valid paths are filtered by determining whether the paths between nodes traverse the path obstacles. The filtering logic has a clear target orientation. Compared to the traditional probabilistic path graph (PRM) mode that performs point-by-point collision detection during the sampling stage, this method only performs targeted verification on the connection paths between skeleton nodes, avoiding repeated detection of the nodes themselves. At the same time, by directly deleting invalid paths that traverse obstacles, it is ensured that all path edges in the final generated probabilistic path graph are within the feasible region, fundamentally eliminating the risk of planned paths colliding with obstacles. Notably, the boundaries of the obstacle map can be imported simultaneously when importing the path obstacles.

[0042] In some embodiments, before connecting any two skeleton nodes to obtain a global path graph including all paths between nodes, the method further includes: Obtain the starting point and the ending point of the path; The starting point and the ending point of the path are defined as skeleton nodes.

[0043] It should be noted that directly defining the path start point and the path end point as skeleton nodes ensures that the topology of the global path graph extends from the start point to the end point. This avoids the problem of traditional probabilistic path graphs not covering the start and end points due to random sampling, which requires additional path connection. This allows subsequent path search to be completed directly within the closed topology network without needing to adapt the start and end points outside the path graph. This ensures that the generated path can accurately connect the start and end positions of the task, improving the completeness of path planning.

[0044] In some embodiments, the step of searching the collision-free probability roadmap based on the shortest path priority principle to obtain the shortest target path from the path start point to the path end point includes: The graph search algorithm is used to search for multiple collision-free feasible paths from the starting point of the path to the ending point of the path, and the collision-free path length of the multiple collision-free feasible paths is calculated. Based on the shortest path priority principle, multiple collision-free path lengths are selected to obtain the shortest path length; Search for the collision-free feasible path corresponding to the shortest path length to obtain the shortest target path.

[0045] It should be noted that, as Figure 16 , 17 As shown in Figure 18, the graph search algorithm traverses the collision-free probability roadmap to generate multiple collision-free feasible paths. Then, based on the length of the collision-free paths, quantitative filtering is performed to ensure that the final output target path strictly meets the core requirement of "shortest path priority". At the same time, since the collision-free probability roadmap has pre-filtered invalid path edges that cross obstacles, all candidate feasible paths have natural collision-free attributes, and there is no need to perform additional collision detection in the search stage, which ensures the safe passage of the path from the underlying logic.

[0046] It should be noted that the vertices of the collision-free probabilistic roadmap are key environmental skeleton nodes (including start and end points), and its topological scale is much smaller than the random sampling topology of traditional probabilistic roadmaps. When performing multi-path search based on this simplified topology, the number of traversed nodes and path edges of the graph search algorithm is greatly reduced, effectively reducing the complexity of path length calculation. At the same time, the process of "generating candidate paths first and then selecting the shortest path" avoids the local optimum problem that may occur in traditional single-path search. While ensuring global optimality, it significantly improves the real-time performance of the search, which is especially suitable for large-scale map scenarios.

[0047] It should be noted that, for example, Dijkstra's algorithm for graph search is used for path search, assuming the probabilistic path graph diagram is as follows: Figure 19As shown, A is the starting point and D is the ending point. The loss (distance) between nodes is marked on the path. The process is as follows: 1) Starting from A, the distance from A to C is 2, and the distance from A to B is 9. Save these to the list {{A, 0, reachable}, {B, 9}, {C, 2}, {D, inf}}. Select the path AC with the smallest loss of 2 from the list. The next path starts from node C, and the list is updated to {{A, 0, reachable}, {B, 9}, {C, 2, reachable}, {D, inf}}; 2) Starting from C, the distance from C to D is 11, and the distance between ACD is 2 + 11 = 13. Save these. 1) In the list {{A, 0, reachable}, {B, 9}, {C, 2, reachable}, {D, 13, reachable}}, select the path AB with the minimum loss of 9. The next path starts from node B, and the list is updated to {{A, 0, reachable}, {B, 9, reachable}, {C, 2, reachable}, {D, 13, reachable}}; 2) Starting from B, the distance between ABD is 9+1=10<13, and the list is updated to {{A, 0, reachable}, {B, 9, reachable}, {C, 2, reachable}, {D, 10, reachable}}; the destination is found and there are no unreachable nodes in the list; 3) The optimal path is ABD.

[0048] Secondly, such as Figure 20 As shown, this application provides a path planning system based on skeleton nodes and probabilistic road maps, applied to the path planning method based on skeleton nodes and probabilistic road maps as described in any one of the first aspects, including: Map acquisition module 10 is used to acquire an obstacle map including path obstacles; The expansion operation module 20 is used to expand the edges of the obstacle map and the path obstacles to obtain the map skeleton. The node extraction module 30 is used to extract the line intersections in the map skeleton to obtain the skeleton nodes of the map skeleton. Road map construction module 40 is used to construct a collision-free probability road map for obstacle avoidance based on the skeleton nodes; The path filtering module 50 is used to search the collision-free probability road map based on the shortest path priority principle to obtain the shortest target path from the path start point to the path end point.

[0049] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the path planning method based on skeleton nodes and probabilistic road maps as described in any one of the first aspects.

[0050] The device / system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0051] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0052] It should be understood that the terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms “a,” “an,” and “described” as used herein may also mean including the plural forms. The terms “comprising,” “including,” “containing,” and “having” are inclusive and therefore indicate the presence of stated features, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof. The method steps, processes, and operations described herein are not construed as requiring them to be performed in a specific order described or illustrated unless the order of execution is explicitly indicated. It should also be understood that additional or alternative steps may be used. The above description is merely a specific embodiment of the invention to enable those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A path planning method based on skeleton nodes and probabilistic road maps, characterized in that, include: Obtain an obstacle map that includes path obstacles; The edges of the obstacle map and the path obstacles are expanded to obtain the map skeleton; Extract the line intersections from the map skeleton to obtain the skeleton nodes of the map skeleton; A collision-free probability roadmap for obstacle avoidance is constructed based on the skeleton nodes; Based on the shortest path priority principle, the collision-free probability roadmap is searched to obtain the shortest target path from the path start point to the path end point.

2. The path planning method based on skeleton nodes and probabilistic road maps according to claim 1, characterized in that, The process of expanding the edges of the obstacle map and the edges of the obstacles to obtain the map skeleton includes: Convert the obstacle map into an obstacle grid map; Define the edges of the obstacle grid map as additional obstacles, and perform an expansion operation on the additional obstacles and the path obstacles; Detect the expanded distance between the additional obstacle and the path obstacle after the expansion operation is performed; If the distance after dilation is equal to the grid number threshold, then the dilation operation is stopped and the map skeleton is obtained.

3. The path planning method based on skeleton nodes and probabilistic road maps according to claim 2, characterized in that, The process of converting the obstacle map into an obstacle grid map includes: Set the conversion resolution for converting the obstacle map into a raster map; The obstacle map is converted into the obstacle grid map based on the conversion resolution.

4. The path planning method based on skeleton nodes and probabilistic road maps according to claim 1, characterized in that, The step of extracting line intersections from the map skeleton to obtain the skeleton nodes of the map skeleton includes: The map skeleton is simplified into a line skeleton map; Extract the line intersections from the line skeleton map to obtain the skeleton nodes of the map skeleton.

5. The path planning method based on skeleton nodes and probabilistic road maps according to claim 1, characterized in that, The construction of a collision-free probability roadmap for obstacle avoidance based on the skeleton nodes includes: Connect any two of the skeleton nodes to obtain a global path graph that includes all paths between nodes; The global path map is filtered based on the path obstacles to obtain the collision-free probability road map.

6. The path planning method based on skeleton nodes and probabilistic road maps according to claim 5, characterized in that, The step of filtering the global path graph based on the path obstacles to obtain the collision-free probability road graph includes: Import the path obstacles into the global path graph; Determine whether the path between the nodes crosses any obstacles along the path; If the path between nodes crosses the path obstacle, then the path between nodes is deleted to obtain the collision-free probability roadmap.

7. The path planning method based on skeleton nodes and probabilistic road maps according to claim 5, characterized in that, Before connecting any two of the skeleton nodes to obtain a global path graph including all paths between nodes, the method further includes: Obtain the starting point and the ending point of the path; The starting point and the ending point of the path are defined as skeleton nodes.

8. The path planning method based on skeleton nodes and probabilistic road maps according to any one of claims 1-7, characterized in that, The process of searching the collision-free probability roadmap based on the shortest path priority principle to obtain the shortest target path from the path start point to the path end point includes: The graph search algorithm is used to search for multiple collision-free feasible paths from the starting point of the path to the ending point of the path, and the collision-free path length of the multiple collision-free feasible paths is calculated. Based on the shortest path priority principle, multiple collision-free path lengths are selected to obtain the shortest path length; Search for the collision-free feasible path corresponding to the shortest path length to obtain the shortest target path.

9. A path planning system based on skeleton nodes and probabilistic road maps, characterized in that, The path planning method based on skeleton nodes and probabilistic road maps as described in any one of claims 1-8, comprising: The map acquisition module is used to acquire an obstacle map, including path obstacles; The expansion module is used to expand the edges of the obstacle map and the path obstacles to obtain the map skeleton. The node extraction module is used to extract the line intersections in the map skeleton to obtain the skeleton nodes of the map skeleton. The road map construction module is used to construct a collision-free probability road map for obstacle avoidance based on the skeleton nodes; The path filtering module is used to search the collision-free probability roadmap based on the shortest path priority principle to obtain the shortest target path from the path start point to the path end point.

10. A computer device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the path planning method based on skeleton nodes and probabilistic road maps as described in any one of claims 1-8.

Images