Robot path planning method and system based on path position point-by-point tracking

Abstract

The invention provides a robot path planning method and system based on path position point-by-point tracking, and belongs to the technical field of robot path planning. Comprising the steps of performing safe spatial domain coverage on an obstacle, and generating an intermediate node; searching the end point by adopting a node drop point detection strategy and an intermediate node search strategy to obtain a path connecting the starting point and the end point; deleting redundant points from the path to generate an initial path; segmenting the initial path to obtain a position coordinate point set; and point-by-point tracking is carried out on the coordinate points by adopting an APF algorithm so as to obtain a final path. According to the method, the problems of high randomness, low sampling efficiency, node redundancy and the like in the robot path planning process can be solved, and the calculation efficiency and the environment adaptability in the robot path planning process are considered.

Description

Technical Field

[0001] The present invention belongs to the technical field of robot path planning, and in particular relates to a robot path planning method and system based on point-by-point tracking of path positions. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The level of development in robotics technology is a key indicator of industrial technological strength. With the rapid evolution of modern science and technology, robotics has achieved widespread application across multiple fields and dimensions. As a key branch of robotics, mobile robots, with their exceptional environmental adaptability and mission-enabling capabilities, have demonstrated significant practical value in key areas such as security, emergency medical rescue, intelligent logistics and distribution, urban environmental maintenance, and smart home services. In mobile robot autonomous navigation systems, path planning is a core module, and its performance directly determines the robot's operational efficiency and reliability.

[0004] Among the current mainstream path planning algorithms, rapidly expanding random trees (RRT) and artificial potential fields (APF) are widely used due to their respective advantages. However, robot path planning based on these path planning algorithms faces some difficult-to-solve technical issues, such as: (1) Due to its inherent random sampling mechanism, the standard RRT algorithm suffers from weak goal orientation when planning paths in complex environments (e.g., narrow passages and dense obstacle areas). This directly leads to low planning efficiency and poor path quality. These problems severely restrict the application of mobile robots in scenarios requiring rapid response or efficient operation (e.g., logistics and rescue).

[0005] (2) The standard APF algorithm is prone to falling into local extreme value traps (such as U-shaped obstacle areas) during path planning, which can cause the robot to be unable to escape and planning to fail. More importantly, in specific scenarios where the end point is close to an obstacle, this type of algorithm has an inherent "unreachable target" problem, that is, the robot cannot ultimately reach the target point or oscillates and stagnates near the target point due to the imbalance between the repulsive force of the obstacle and the gravitational force of the end point. This greatly limits the robustness and practicality of the algorithm in complex and dynamic environments, making the planned path unable to reliably guide the robot to complete the task. Summary of the Invention

[0006] In order to overcome the shortcomings of the above-mentioned existing technologies, the present invention provides a robot path planning method and system based on point-by-point tracking of path positions, which can solve the problems of strong randomness, low sampling efficiency and node redundancy in the robot path planning process, and thus take into account the computational efficiency and environmental adaptability of the robot path planning.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: A first aspect of the present invention provides a robot path planning method based on point-by-point tracking of path positions.

[0008] A robot path planning method based on point-by-point path position tracking, comprising: Construct a grid map and determine the location of obstacles; cover the obstacles with a safe spatial domain and generate intermediate nodes; Based on the obtained intermediate nodes, the node landing detection strategy and the intermediate node search strategy are used to search for the end point to obtain a path connecting the starting point and the end point; redundant points are deleted from the obtained path to generate an initial path; The initial path is segmented according to the set step size to obtain the position coordinate point set of the initial path; The APF algorithm is used to track the coordinate points in the obtained position coordinate point set point by point, and the path formed by point-by-point tracking is used as the final path.

[0009] Furthermore, when the obstacle is covered in the safety space, the side length of the covered obstacle is expressed as: ; in, is the short side of the rectangle; is the long side of the rectangle; is the distance from the short side of the rectangle to the safe space domain; is the distance from the long side of the rectangle to the safe space domain; is the step size of the RRT search.

[0010] Furthermore, generating an intermediate node includes: An intermediate node is set at the midpoint of the line connecting the starting point and the end point after system initialization; if the intermediate node coincides with an obstacle, a circle is drawn outward with the intermediate node as the center until it is tangent to the boundary of the safe space domain, and the tangent point is updated to the new intermediate node.

[0011] Furthermore, a node landing detection strategy is used for searching, including: Random sampling is performed within the range of the map obstacle information, and the nearest neighbor node of the random sampling point is found in the tree set; the step size is extended along the direction from the nearest neighbor node to the random sampling point to generate a new node; if the generated new node is not within the range of the map obstacle information, it is added to the tree set, otherwise, re-sampling is performed.

[0012] Furthermore, an intermediate node search strategy is used for searching, including: First, a forward path search is performed from the starting point to the intermediate node; then, a backward path search is performed from the intermediate node to the end point. During the forward and backward path searches, it is respectively detected whether the current tree node can be directly connected to the intermediate node and the end point. If feasible, they are directly connected.

[0013] Furthermore, the APF algorithm includes: introducing a first distance factor and a second distance factor into the gravitational potential field and the repulsive potential field in the APF algorithm respectively, so as to obtain improved gravitational function and repulsive function.

[0014] Furthermore, the APF algorithm is used to track the coordinate points in the position coordinate point set point by point, including: When the APF algorithm is used to track the coordinate points in the position coordinate point set, when the target area of any local node is reached, the next node is tracked; this process is repeated to obtain the final path.

[0015] A second aspect of the present invention provides a robot path planning system based on point-by-point tracking of path positions.

[0016] A robot path planning system based on point-by-point path position tracking, comprising: The intermediate node generation module is configured to: construct a grid map and determine the location of obstacles; perform safety space domain coverage on obstacles and generate intermediate nodes; The initial path search module is configured to: search for the end point based on the obtained intermediate nodes using a node landing detection strategy and an intermediate node search strategy to obtain a path connecting the start point and the end point; and remove redundant points from the obtained path to generate an initial path; The initial path segmentation module is configured to: segment the initial path according to the set step size to obtain a set of position coordinate points of the initial path; The path planning module is configured to: use the APF algorithm to track the coordinate points in the obtained position coordinate point set point by point, and use the path formed by the point-by-point tracking as the final path. A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps of a robot path planning method based on point-by-point tracking of path positions as described in the first aspect of the present invention.

[0017] The fourth aspect of the present invention provides an electronic device, comprising a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps of a robot path planning method based on point-by-point tracking of path positions as described in the first aspect of the present invention are implemented.

[0018] One or more of the above technical solutions have the following beneficial effects: (1) The present invention covers the obstacle in a safe spatial domain and generates intermediate nodes. Based on the intermediate nodes obtained, the node landing point detection strategy and the intermediate node search strategy are used to search for the end point to obtain a path connecting the starting point and the end point. Redundant points are then deleted from the obtained path to generate an initial path. Through this design, only the node coordinates need to be judged instead of the line segment collision, which can greatly reduce the amount of collision detection calculations and reduce the planning time in dense obstacle environments to meet the real-time requirements of logistics, rescue and other scenarios. By removing redundant inflection points through the reverse tree node optimization strategy, the path length can be shortened and a smooth initial path closer to the theoretical optimum can be generated to improve the robot's operating efficiency. Therefore, compared with the existing technology, the present invention can significantly improve the efficiency of path planning and the quality of generated paths.

[0019] (2) The present invention introduces a first distance factor and a second distance factor into the gravitational potential field and the repulsive potential field in the APF algorithm, respectively, to obtain improved gravitational function and repulsive function. This can prevent the repulsive force of obstacles near the target from being too large, resulting in the inability to reach the destination, and can also prevent the target point from oscillating or stagnating due to the repulsive force of obstacles. At the same time, the initial path is divided into a sequence of coordinate points, and the improved APF algorithm is used to generate gravitational force for each local node, which can guide the robot to track the nodes in turn and finally generate a smooth path. Therefore, compared with the existing technology, the present invention significantly reduces the running time and path length during path planning in dynamic and mixed obstacle environments, and will not fall into local optimality. Even in specific environments, there will be no problem of unreachable targets.

[0020] Advantages of additional aspects of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0022] Figure 1 This is a flow chart of a robot path planning method based on point-by-point path position tracking in Example 1 of the present invention.

[0023] Figure 2 Schematic diagram of local node gravity tracked point by point in embodiment 1 of the present invention.

[0024] Figure 3 Schematic diagram of the path of the standard RRT algorithm in the prior art.

[0025] Figure 4 Schematic diagram of the path of the RRT-connect algorithm in the prior art.

[0026] Figure 5 Schematic diagram of the path of the standard RRT* algorithm in the prior art.

[0027] Figure 6 This is a schematic diagram of the improved RRT algorithm path in Example 1 of the present invention.

[0028] Figure 7 This is a simulation comparison path diagram of the standard APF algorithm and the improved APF algorithm in a static obstacle environment in Example 1 of the present invention.

[0029] Figure 8 This is a simulation comparison path diagram of the standard APF algorithm and the improved APF algorithm in a dynamic obstacle environment in Example 1 of the present invention.

[0030] Figure 9 This is a simulation comparison path diagram of the standard APF algorithm and the improved APF algorithm in a static and dynamic mixed obstacle environment in Example 1 of the present invention.

[0031] Figure 10 This is a schematic diagram of path tracking under the fusion of the improved APF algorithm and the improved RRT algorithm in Example 1 of the present invention.

[0032] Figure 11 for Figure 10 Winning bid number A partial enlarged schematic diagram of the area.

[0033] Figure 12 for Figure 10 Winning bid number A partial enlarged schematic diagram of the area. DETAILED DESCRIPTION

[0034] It should be noted that the following detailed descriptions are exemplary and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0035] It should be noted that the terms used herein are for describing particular embodiments only and are not intended to limit the exemplary embodiments according to the present invention.

[0036] In the absence of conflict, the embodiments of the present invention and the features thereof may be combined with each other.

[0037] Example 1 This embodiment discloses a robot path planning method based on point-by-point tracking of path positions.

[0038] like Figure 1 As shown, a robot path planning method based on point-by-point tracking of path positions includes: Step S1, system initialization; Step S2: construct a grid map and determine the location of obstacles; perform safety space domain coverage on obstacles after system initialization and generate intermediate nodes; Step S3: Based on the obtained intermediate nodes, the node landing detection strategy and the intermediate node search strategy are used to search for the end point to obtain a path connecting the start point and the end point; redundant points are deleted from the obtained path to generate an initial path; Step S4: Segment the initial path according to the set step size to obtain a set of position coordinate points of the initial path; Step S5: Using the improved APF algorithm to track the coordinate points in the obtained position coordinate point set point by point, and taking the path formed by the point-by-point tracking as the final path.

[0039] Based on the above process, the present invention can solve the problems of strong randomness, low sampling efficiency, and node redundancy in the robot path planning process, thereby balancing computational efficiency and environmental adaptability during robot path planning. To facilitate understanding of the technical solution of the present invention, the specific implementation method of the technical solution of the present invention is further explained and illustrated below.

[0040] In step S1, the system is initialized, including initializing: tree set, map obstacle information, setting the maximum number of iterations, setting the iteration termination condition, setting the starting point coordinates and the end point coordinates, etc.

[0041] In this embodiment, system initialization includes initializing the tree set, map obstacle information, setting the maximum number of iterations, setting the iteration termination condition, setting the starting point coordinates and the end point coordinates, etc. Specifically: A. Initialize the tree set: In path planning, a tree is a data structure used to represent the state and information from the starting point to the current path. Initializing the tree set means starting to build this data structure to provide an initial state for subsequent path expansion and tree construction.

[0042] B. Initialize map obstacle information: This refers to reading the obstacle information data set in the environment and converting it into a raster map suitable for this algorithm to ensure the correct operation of the algorithm.

[0043] C. Set the maximum number of iterations: In order to control the running time and computing resources of the algorithm, it is necessary to set the maximum number of iterations as an upper limit condition for the termination of the algorithm, thereby ensuring that the algorithm does not run indefinitely or occupy too many resources.

[0044] D. Set the iteration termination condition: The condition for iteration termination is whether the maximum number of iterations is reached. If the number of iterations is greater than the set maximum number of iterations, the iteration is terminated and the path search fails.

[0045] E. Set the start and end point coordinates: The start and end points are the basic inputs to the path planning algorithm. The coordinates of these two points must be determined during initialization. By initializing the start and end point coordinates, you can clearly define the area within which the path is planned.

[0046] During the implementation, you can perform the following operations: 1) Start point reachability check: During the initialization phase, first ensure that the start point is reachable in the environment, i.e., it does not overlap with obstacles. If the start point is unreachable, it is necessary to adjust the obstacle model information or reconfigure the initial conditions of the start point.

[0047] 2) Initialize the RRT algorithm: take the starting point as the root node of the tree, and the subsequent algorithm will perform path expansion based on this tree node.

[0048] 3) Read obstacle data: Obtain the location information of obstacles in the environment based on the provided map file.

[0049] 4) Set the tree collection and path: In the initial state, the tree collection only contains the starting point and the path is empty.

[0050] In step S2, the obstacles initialized by the system are covered in a safe space domain and intermediate nodes are generated.

[0051] Because path generation based on sampling strategies can result in line collisions, this invention implements a safe region around obstacles. This allows the algorithm to use the safe region as the obstacle boundary during path exploration, effectively avoiding collisions between the path and the original obstacle. Furthermore, this safe region effectively reduces the risk of collisions between the mobile robot and obstacles during actual path planning.

[0052] When covering the obstacle in the safety space, the side length of the covered obstacle is expressed as: ; in, is the short side of the rectangle; is the long side of the rectangle; is the distance from the short side of the rectangle to the safe space domain; is the distance from the long side of the rectangle to the safe space domain; is the step size of the RRT search.

[0053] Calculated from the above formula and Then, the safety space rectangle is expanded outwards to a certain extent with the original obstacle rectangle as the center, and the safety space domain is covered for the current obstacle. The specific expansion range is: if the length and width of the original obstacle rectangle are 、 , then the length and width of the safety space rectangle are 、 .

[0054] After the safe space domain is fully covered for obstacles, an intermediate node is selected. Specifically, an intermediate node is set at the midpoint of the line connecting the starting and ending points after system initialization. If the intermediate node coincides with an obstacle, a circle is drawn outward from the intermediate node until it intersects the boundary of the safe space domain, and the tangent point is updated to the new intermediate node. Selecting intermediate nodes in this way provides security and stability. Intermediate nodes are generated outside the covered safe space domain, ensuring that the path maintains a safe distance from obstacles throughout the entire process, avoiding direct contact with the actual obstacle boundary, thereby reducing the risk of physical collision. Furthermore, intermediate nodes are obtained by connecting midpoints or expanding circles outward. The selection does not rely on random sampling, but is determined based on obstacle information. This method helps maintain the dynamic adjustment capability of the path when adjusting obstacle information, improving the robustness of the algorithm search process.

[0055] In step S3, based on the obtained intermediate nodes, the node landing detection strategy and the intermediate node search strategy are used to search for the end point to obtain a path connecting the start point and the end point; redundant points are deleted from the obtained path to generate an initial path.

[0056] The search method of sampling strategy in the prior art is generally: random sampling within the map range , find its nearest neighbor node in the tree set ,along arrive Direction expansion step , generate a new node ; If the line does not collide with the obstacle, then connect and It is then added to the tree collection, otherwise it is resampled. Obstacle collision in this search method is determined by whether the connecting line collides with the obstacle boundary. In actual implementation, this requires traversing to check whether the line segment between the new node and the tree node collides with the edge of a rectangle. If there is no collision, it is retained. Therefore, this method is highly complex and inefficient in environments with many obstacles on the map.

[0057] To this end, the present invention adopts a node landing detection strategy for search, including: Random sampling within the range of map obstacle information , and find the nearest neighbor node of the randomly sampled point in the tree set ; Along the nearest neighbor node To random sampling point The direction expansion step , to generate a new node ; If the new node generated If it is not within the range of the map obstacle information, it will be added to the tree set, otherwise, it will be re-sampled. Compared with the existing technology, the present invention adopts the landing point detection strategy, and each time a new node is obtained When the node is found, it checks whether it is in free space outside of map obstacles. If so, the new node is retained. In practice, selecting a new node is accomplished simply by determining whether the new node's coordinates are in free space, eliminating the need for traversal queries. This significantly improves search efficiency, especially in complex environments with many obstacles.

[0058] In addition, the present invention also adopts an intermediate node search strategy to perform the search, including: First, a forward path search is performed from the starting point to the intermediate node; then, a backward path search is performed from the intermediate node to the end point. During both the forward and backward path searches, the current tree node is checked to see if it can be directly connected to the intermediate node and the end point. If feasible, a direct connection is made. This, on the one hand, can accelerate search efficiency by breaking the path between the starting point and the end point into two segments and prioritizing paths that can be directly connected, thereby quickly reaching the intermediate node and the end point. This reduces the number of samples, speeds up the search process, and avoids the resource waste caused by blind expansion in existing technologies. On the other hand, it can reduce computational complexity by breaking the entire path planning into two parts, processing each part separately, and reducing the computational burden.

[0059] Through searching, a path connecting the starting point and the end point can be obtained; then, the reverse tree node optimization strategy can be used to delete redundant points from the obtained path to generate an initial path. Specifically, the reverse tree node optimization strategy is: using the greedy idea, starting from the end point to the starting point, perform tree node traversal queries on the generated path, and detect the collision relationship between the connection line and the obstacle with the current tree node. If the connection line does not collide with the obstacle, check whether the next point collides with the obstacle; if the connection line with the next tree node is detected to collide with the obstacle, the current tree node is used as the final connection line. Then, repeat the above operation with the current tree node until it is connected to the starting point; finally, the initial path is output. Under this strategy, the reverse tree node optimization example is as follows: Figure 6 shown.

[0060] In step S4, the generated initial path is segmented according to the set step size to obtain a set of position coordinate points of the initial path.

[0061] First, the node-optimized route is set as the initial path, and the coordinates of all turning points including the starting point and the end point are recorded.

[0062] Then, set the split step size , The closer to the step size of the RRT algorithm The larger the offset planned by the algorithm is and the closer it is to 0, the closer the path is to the initial path; among them, the segmentation step length The value can be set according to actual situation.

[0063] Then, starting from the starting point, split the step The next inflection point is divided in sequence for the spacing until a certain segmentation step exceeds the inflection point, thereby obtaining the coordinate point set of the first segment; the same operation is performed at the next inflection point until the segmentation reaches the end point.

[0064] Finally, the coordinate point sets of each segment are merged in sequence to obtain the initial position coordinate point set.

[0065] In step S5, the APF algorithm is used to track the coordinate points in the obtained position coordinate point set point by point, and the path formed by the point-by-point tracking is used as the final path.

[0066] The first distance factor and the second distance factor are introduced into the gravitational potential field and the repulsive potential field in the APF algorithm respectively to obtain the improved gravitational function and repulsive function. Specifically: A. Introducing a distance factor into the gravitational potential field , which is the first distance factor. The improved gravity function is as follows: ; in, is the gravitational gain coefficient of the end point; is the vector distance from the robot to the target point, is the current position of the robot, is the end position; is the critical distance, within which the traditional linear gravity formula is used; To adjust the parameters for the smooth transition of gravity changes at long distances, which is related to the map size; To control the smoothing parameter, the smaller the value, the smoother the force field.

[0067] By introducing the first distance factor in the gravitational potential field: 1) When the robot is farther from the target point than the critical distance When the gravity function gradually reduces the growth rate of gravity through a smooth transition (such as the sigmoid function). This design avoids the path redundancy or jitter that may be caused by linear gravity in the standard APF algorithm at long distances, that is, the gravity strength can be dynamically adjusted. 2) At the critical distance Nearby, the gravity function achieves continuous transition through the smooth characteristics of the exponential function, reducing the sudden change of force, thereby reducing the risk of oscillation caused by drastic changes in gravity when the mobile robot approaches the target, and thus avoiding sudden changes in gravity.

[0068] B. Introducing a distance factor into the repulsive potential field , which is the second distance factor. The improved repulsion function As shown below: ; ; in, is the repulsion gain coefficient of the obstacle; is the vector distance from the obstacle to the robot, is the location of the obstacle; is a constant, indicating the maximum distance at which an obstacle can exert repulsive force on the robot; is the target threshold distance. When the distance between the robot and the target point is less than this value, the repulsive force begins to decrease.

[0069] By introducing a second distance factor in the repulsive potential field: 1) When the mobile robot approaches the target (the distance is less than the target threshold distance When the second distance factor The repulsive force of the obstacle is inversely proportional to the distance between the mobile robot and the target, which gradually weakens the repulsive force. Even if there is an obstacle near the target, the repulsive force will not completely offset the gravitational force, thus ensuring that the mobile robot can reach the destination. In other words, the repulsive force near the target can be dynamically suppressed. 2) By limiting the maximum effective distance of the repulsive force , and combined with the second distance factor The dynamic adjustment of avoids the problem of the obstacle repulsion having a continuous impact on the global scope in the standard APF algorithm, reduces the generation of local extreme values, and thus can adapt the repulsion range.

[0070] At the same time, the first distance factor and the second distance factor also have a synergistic effect, that is, near the target, the improved gravitational function dominates the path tracking, while the repulsive function actively retreats. The two work together to avoid the antagonistic balance of forces, thereby eliminating oscillation or stagnation.

[0071] Therefore, by introducing a distance factor, the improved APF algorithm achieves a dynamic weighting of attraction and repulsion: as the robot moves away from the target, the attraction gradually increases to improve guidance efficiency; as the robot approaches the target, the repulsion actively decreases to avoid interfering with the destination. Furthermore, the smooth transition of forces and the optimization of local action ranges reduce oscillations and redundancy in path planning, ultimately resolving the target unreachability and local extrema issues of the standard APF algorithm in the prior art.

[0072] The improved APF algorithm of the present invention is used to track the coordinate points in the position coordinate point set point by point, including: The generated initial path is divided into several coordinate points according to a certain step size. Then, the improved APF algorithm is used to track the points one by one. The coordinate points within a certain range will generate gravity for the APF algorithm. When the target area of a local node is reached, the next node will be tracked, and the final path will be obtained. The schematic diagram of the local node gravity of point-by-point tracking is shown in the figure below. Figure 2 As shown, and are the gravitational force of the target point and the repulsive force of the obstacle, is the resultant force of the target point and the obstacle; and is the node gravity, is the resultant force of the node’s gravity; is the total force. Furthermore, the calculation formula of the local node gravity is as follows: ; in, For the The gravitational force of a local node; is the gain coefficient; For robots To local nodes distance; is the local gravitational range.

[0073] To further demonstrate the advantages of the method provided by the present invention, this embodiment performs the following simulation comparisons, specifically: 1) Compare the paths generated by the improved RRT algorithm of the present invention with those generated by the standard RRT algorithm, the RRT-connect algorithm, and the standard RRT* algorithm.

[0074] The simulation results are shown in the figure Figure 3-Figure 6 As shown, Figure 3 Schematic diagram of the path of the standard RRT algorithm in the prior art, Figure 4 A schematic diagram of the path of the RRT-connect algorithm in the prior art, Figure 5Schematic diagram of the path of the standard RRT* algorithm in the prior art; Figure 6 The figure is a schematic diagram of the improved RRT algorithm path of the present invention. Figure 3-Figure 6 It can be seen from the figure that the improved RRT algorithm of the present invention has fewer redundant points and the shortest path compared with other algorithms.

[0075] In this embodiment, each set of experiments is repeated 1000 times, and the path length, running time, and sampling times are averaged to obtain the final data, which is shown in Table 1.

[0076] Table 1 Data comparison of different algorithms in the same map environment

[0077] As can be seen from Table 1, in this map environment ( Figure 3-Figure 6 In the same map environment, the improved RRT algorithm of this invention reduced runtime by 89.33%, 61.29%, and 94.98% compared to the standard RRT algorithm, RRT-connect, and standard RRT* algorithms, respectively. It also reduced path length by 18.51%, 12.83%, and 1.20%, and the number of samples by 85.77%, 29.76%, and 82.47%, respectively. This data demonstrates that the improved RRT algorithm can effectively reduce runtime, improve convergence speed, and shorten path length.

[0078] 2) The improved APF algorithm of the present invention is simulated and verified.

[0079] Three environments are set up, namely static obstacle environment, dynamic obstacle environment and static and dynamic mixed obstacle environment. The simulation results are as follows Figure 7-Figure 9 As shown, Figure 7 The simulation comparison path diagram of the standard APF algorithm and the improved APF algorithm of the present invention in a static obstacle environment is shown in FIG. Figure 8 This is a simulation comparison path diagram of the standard APF algorithm and the improved APF algorithm of the present invention in a dynamic obstacle environment. Figure 9 The figure is a simulation comparison path diagram of the standard APF algorithm and the improved APF algorithm of the present invention under static and dynamic mixed obstacle environment. Figure 7-Figure 9 In the figure, a green triangle represents the starting point, a magenta triangle represents the ending point, a gray circle represents a static obstacle, and a magenta circle represents a dynamic obstacle. A thin dotted line represents the trajectory of the movement, a blue dashed line represents the planned path of the standard APF algorithm, and a red dashed line represents the planned path of the improved APF algorithm. The data comparison is shown in Table 2.

[0080] Table 2 Data comparison of different algorithms under three environments

[0081] Depend on Figure 7 Static map simulation shows that the standard APF algorithm is unable to reach the target point, while the improved algorithm improves the repulsive potential field and can successfully reach the destination even when there are obstacles near the destination. As shown in Table 2, in a dynamic obstacle environment and a static and dynamic mixed environment, the improved algorithm of the present invention shortens the running time by 35.71% and 41.94% respectively compared to the standard APF algorithm, and shortens the number of running steps and path length by 2.03% and 6.43% overall. Therefore, the improved APF algorithm of the present invention has better results in route planning than the traditional APF algorithm.

[0082] 3) The improved RRT algorithm and APF algorithm are integrated.

[0083] like Figure 10 As shown, the improved RRT algorithm and APF algorithm of the present invention are fused; Figure 11 and Figure 12 They are the fusion simulation results ( Figure 10 ) Winning bid number and label A magnified view of the indicated area. Figure 10-12 In the figure, the green triangle represents the starting position, the magenta triangle represents the end position, the black rectangle represents the obstacle, the red dotted line outside the obstacle represents the covered safe space, the red solid line is the initial path planned by the improved RRT algorithm of the present invention, and the blue solid line is the path planned by the fusion of the improved RRT and improved APF algorithms. Figure 11 and Figure 12 It can be seen from the figure that the route finally planned by the robot path planning method provided by the present invention is smoother and has no inflection points, which can significantly improve the stability of the robot during movement.

[0084] Example 2 This embodiment discloses a robot path planning system based on point-by-point tracking of path positions.

[0085] A robot path planning system based on point-by-point path position tracking, comprising: The initialization module is configured to: perform system initialization; The intermediate node generation module is configured to: construct a grid map and determine the location of obstacles; perform safety space domain coverage on obstacles initialized by the system and generate intermediate nodes; The initial path search module is configured to: search for the end point based on the obtained intermediate nodes using a node landing detection strategy and an intermediate node search strategy to obtain a path connecting the start point and the end point; and remove redundant points from the obtained path to generate an initial path; The initial path segmentation module is configured to: segment the initial path according to the set step size to obtain a set of position coordinate points of the initial path; The path planning module is configured to: use the improved APF algorithm to track the coordinate points in the obtained position coordinate point set point by point, and use the path formed by the point-by-point tracking as the final path. Example 3 The purpose of this embodiment is to provide a computer-readable storage medium.

[0086] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a robot path planning method based on point-by-point path position tracking as described in the first embodiment of the present disclosure.

[0087] Example 4 The purpose of this embodiment is to provide an electronic device.

[0088] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps of a robot path planning method based on point-by-point path position tracking as described in the first embodiment of the present disclosure are implemented.

[0089] The steps involved in the apparatuses of Examples 2, 3, and 4 above correspond to those of Method Example 1. For detailed implementations, please refer to the relevant description of Example 1. The term "computer-readable storage medium" should be understood to mean a single medium or multiple media containing one or more instruction sets; it should also be understood to include any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and causing the processor to perform any method of the present invention.

[0090] Those skilled in the art will appreciate that the modules or steps of the present invention described above can be implemented using a general-purpose computer device. Alternatively, they can be implemented using program code executable by a computing device, which can then be stored in a storage device and executed by the computing device. Alternatively, they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0091] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A robot path planning method based on point-by-point tracking of path positions, characterized in that: include: Build a grid map and determine the location of obstacles; Cover the obstacles in a safe space and generate intermediate nodes; Based on the obtained intermediate nodes, the node landing detection strategy and the intermediate node search strategy are used to search for the end point to obtain a path connecting the starting point and the end point; redundant points are deleted from the obtained path to generate an initial path; Segment the initial path according to the set step size to obtain a set of position coordinate points of the initial path; The APF algorithm is used to track the coordinate points in the obtained position coordinate point set point by point, and the path formed by point-by-point tracking is used as the final path.

2. A robot path planning method based on point-by-point path position tracking according to claim 1, characterized in that: When covering the obstacle in the safety space, the side length of the covered obstacle is expressed as: ； in, is the short side of the rectangle; is the long side of the rectangle; is the distance from the short side of the rectangle to the safe space domain; is the distance from the long side of the rectangle to the safe space domain; is the step size of the RRT search.

3. A robot path planning method based on point-by-point path position tracking according to claim 1, characterized in that: Generate intermediate nodes, including: An intermediate node is set at the midpoint of the line connecting the starting point and the end point after system initialization; if the intermediate node coincides with an obstacle, a circle is drawn outward with the intermediate node as the center until it is tangent to the boundary of the safe space domain, and the tangent point is updated to the new intermediate node.

4. A robot path planning method based on point-by-point path position tracking according to claim 1, characterized in that: The node drop detection strategy is used for search, including: Random sampling is performed within the range of the map obstacle information, and the nearest neighbor node of the random sampling point is found in the tree set; the step size is extended along the direction from the nearest neighbor node to the random sampling point to generate a new node; if the generated new node is not within the range of the map obstacle information, it is added to the tree set, otherwise, re-sampling is performed.

5. A robot path planning method based on point-by-point path position tracking according to claim 1, characterized in that: The intermediate node search strategy is used for searching, including: First, a forward path search is performed from the starting point to the intermediate node; then, a backward path search is performed from the intermediate node to the end point. During the forward and backward path searches, it is respectively detected whether the current tree node can be directly connected to the intermediate node and the end point. If feasible, they are directly connected.

6. A robot path planning method based on point-by-point path position tracking according to claim 1, characterized in that: The APF algorithm includes: introducing a first distance factor and a second distance factor into the gravitational potential field and the repulsive potential field in the APF algorithm respectively, so as to obtain improved gravitational function and repulsive function.

7. A robot path planning method based on point-by-point path position tracking according to claim 1, characterized in that: The APF algorithm is used to track the coordinate points in the position coordinate point set point by point, including: When the APF algorithm is used to track the coordinate points in the position coordinate point set, when the target area of any local node is reached, the next node is tracked; this process is repeated to obtain the final path.

8. A robot path planning system based on point-by-point tracking of path positions, characterized in that: include: The intermediate node generation module is configured to: construct a grid map and determine the location of obstacles; Cover the obstacles in a safe space and generate intermediate nodes; The initial path search module is configured to: search for the end point based on the obtained intermediate nodes using a node landing detection strategy and an intermediate node search strategy to obtain a path connecting the start point and the end point; and remove redundant points from the obtained path to generate an initial path; The initial path segmentation module is configured to: segment the initial path according to the set step size to obtain a set of position coordinate points of the initial path; The path planning module is configured to: use the APF algorithm to track the coordinate points in the obtained position coordinate point set point by point, and use the path formed by the point-by-point tracking as the final path.

9. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by a processor, the steps of the robot path planning method based on point-by-point path position tracking as described in any one of claims 1 to 7 are implemented.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the robot path planning method based on point-by-point path position tracking as described in any one of claims 1 to 7 are implemented.

Images