A mobile robot path planning method based on spatially constrained A-star algorithm

By introducing the constraints of the robot entity covering the grid matrix S and the safe braking distance L in the A-star algorithm, the problem that the A-star algorithm's planning path does not meet the actual robot size is solved, and the best path search under obstacles is achieved and safely pass.

CN114578828BActive Publication Date: 2024-11-22CHONGQING UNIV OF POSTS & TELECOMM
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210292480.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-23
Publication Date
2024-11-22
Estimated Expiration
2042-03-23

AI Technical Summary

Technical Problem

When planning the path, the A-star algorithm relies on the grid map resolution, resulting in the planned path not meeting the actual passing situation when the actual carrier size of the robot is larger than the map resolution, and there is a problem that it does not conform to the actual robot path.

Method used

The A-star algorithm introduces constraints on the robot entity covering the grid matrix S and the safe braking distance L. Through these constraints, the path selection planning is carried out to ensure the width and safety of the path.

Benefits of technology

It realizes searching for the best route when there are obstacles, can adapt to the location changes of obstacles, and the planned path meets the robot width requirements and ensures safe passage, avoiding the problem of paths being too narrow or collision in traditional A-star algorithms.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114578828B_ABST
    Figure CN114578828B_ABST
Patent Text Reader

Abstract

The present invention discloses a mobile robot path planning method based on a spatially constrained A-star algorithm: the method first constructs a grid map of the robot's working environment, and on this basis marks passable grids and impassable grids, and also introduces robot entity coverage grid matrix S constraints and safety braking distance L constraints into the A-star algorithm, and finally finds the shortest feasible path from the start point to the end point based on the improved A-star algorithm, so that the robot can pass smoothly and reach the target position safely. The present invention can adapt to grid maps of different resolutions, dynamically adjust the robot's entity coverage grid matrix S and safety braking distance L in the grid map, and accurately plan a safe path that meets the machine width.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of mobile robot path planning, and relates to a mobile robot path planning method based on a space-constrained A-star algorithm. Background Art

[0002] Path planning is a hot research topic in intelligent technology, and has made breakthroughs in many fields and has been successfully applied. From traditional algorithms to algorithms developed in combination with bionics, intelligent algorithms have made great progress. Different intelligent algorithms have different characteristics, and their scope and fields of application are also different. Therefore, studying path planning intelligent algorithms from the perspective of the characteristics of the algorithm itself and its application is of great significance to the development of path planning technology.

[0003] The A-star algorithm is a classic heuristic search algorithm that is suitable for global path planning in a known environment. It is optimal, complete, and efficient. However, due to the characteristics of the A-star algorithm itself, the planned path is completely dependent on the grid map resolution, that is, the path width planned by the A-star algorithm is consistent with the map resolution. Therefore, when the actual robot carrier size is larger than the map resolution, the planned path may not match the actual machine path due to the limitations of the A-star algorithm itself. Summary of the invention

[0004] The purpose of the present invention is to provide a method for introducing robot entity coverage grid matrix S constraint and safety braking distance L constraint into the A-star algorithm to accurately plan a safe path that meets the robot width.

[0005] The technical solution adopted to achieve the above purpose is: a mobile robot path planning method based on the spatial constraint A star algorithm, characterized in that it includes the following steps:

[0006] Step 1: Construct a grid map based on the actual work scene area and the required map resolution, and mark the grids occupied by scene obstacles in the map as inaccessible grids;

[0007] Step 2: Combined with the grid map Map established in step 1, determine the robot's physical coverage grid matrix S and safe braking distance L;

[0008] Step 3: In the A-star algorithm, two constraints, the robot entity coverage grid matrix S and the safe braking distance L, are added to perform path optimization planning to obtain a safe and effective path for the robot from the starting grid to the target grid.

[0009] The beneficial effects of the present invention due to the technical solution are as follows:

[0010] (1) The present invention adopts the A-star algorithm, and on this basis introduces the constraints of the robot's physical coverage area and safe braking distance, continuously updates the center position of the robot, and can directly search for the best route. It can also complete the search for the best route in the presence of obstacles.

[0011] (2) The method can be extended to solve practical problems and can complete the search for the best route when the position of obstacles changes. It is intuitive and clear and has strong applicability.

[0012] (3) The planned path meets a certain width and can ensure the smooth passage of the robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a flow chart of the method of the present invention;

[0014] Figure 2 is a schematic diagram of a regional scene in an embodiment of the present invention;

[0015] Figures 3-1 to 3-18 It is a schematic diagram of an embodiment of a mapped grid matrix S' obtained by mapping the grid matrix S of the present invention. DETAILED DESCRIPTION

[0016] See also Figure 1 :A mobile robot path planning method based on spatial constraint A-star algorithm, comprising the following steps:

[0017] Step 1: Construct a grid map based on the actual work scene area and the required map resolution, and mark the grids occupied by scene obstacles in the map as inaccessible grids;

[0018] Step 2: Combined with the grid map Map established in step 1, determine the robot's physical coverage grid matrix S and safe braking distance L;

[0019] Step 3: In the A-star algorithm, two constraints, the robot entity coverage grid matrix S and the safe braking distance L, are added to perform path optimization planning to obtain a safe and effective path for the robot from the starting grid to the target grid.

[0020] In step 1, a grid map Map is constructed according to the actual working scene area and the required map resolution, and the grids occupied by scene obstacles in the Map are marked as inaccessible grids, specifically:

[0021] According to the actual working scene area and the required map resolution m×n, the grid map Map is constructed using the grid method. The working environment is divided into an m×n grid matrix. The length and width of each grid are k (mm). According to the grid position occupied by the scene obstacles in the grid map Map, the inaccessible grids are marked, and the remaining grids are passable grids.

[0022] The step 2 combines the grid map Map established in step 1 to determine the robot's physical coverage grid matrix S and safe braking distance L, specifically:

[0023] The robot's physical coverage grid matrix S is the set of grids occupied by the robot in the grid map Map at the current moment, and the safe braking distance L is the side length k of the grid in the grid map Map.

[0024] The method for optimal path planning described in step 3 is:

[0025] Step 3.1: In the grid map Map, determine the starting grid and target grid of the robot path planning, and use M to represent the center grid of the grid matrix S where the robot entity covers the jth moment. j-1 To represent, where j is an integer (1 to the number of grids in path planning), j-1 represents the number of parent nodes of the central grid, and when j-1=0, there is no parent node grid; during initialization, j=1, the initial central grid M0 coincides with the starting grid, and then the complete information set of the initial central grid M0 is put into the close list to complete the initialization of path planning. The complete information set of the central grid includes: the central grid M j-1 The coordinates of the grid, the parent node grid M (j-1)-1 Coordinates and path cost F; use close list to store the planned paths center The complete information set of the grid. The grid in the close list will no longer participate in the subsequent path planning calculation. The open list is used to store the partial information set of the passable grid in the path planning calculation. The partial information set of the passable grid includes: the coordinates of the grid and the path cost F;

[0026] Step 3.2: Determine the grid M j-1 Adjacent eight-domain grid N i The ith grid in the grid (i is an integer from 1 to 8) is an inaccessible grid, or a grid in the close list. If it is an inaccessible grid, the grid is ignored. i ; If it is a passable grid, proceed to the next step;

[0027] Step 3.3: According to the accessible grid N i The grid matrix S is mapped with the safe braking distance L to obtain the mapped grid matrix S', and then it is determined whether there is an impassable grid in the grid matrix S'. If so, the grid N is ignored. i Then return to step 3.2;

[0028] Step 3.4: Calculate the traversable grid N i The path cost F is used to determine the grid N iIs it in the open list? If not, put the grid N i The current partial information set is put into the open list. If the grid N i If it already exists in the open list, the grid N i The current partial information set replaces the existing partial information set;

[0029] Step 3.5: Determine grid N i Is i less than or equal to 8? If so, return to step 3.2. After traversing all eight domain grids, sort all grids in the open list according to the F value from small to large, and take the grid with the smallest F value as the central grid M of the robot's entity coverage grid matrix S at the next moment. j , and then the grid M j Remove the open list and then move the current center grid M j-1 Set as the next moment grid M j The parent node grid; finally, the grid M j-1 The complete information set is moved to the close list;

[0030] Step 3.6: Repeat steps 3.2, 3.3, 3.4 and 3.5 until the target grid is added to the closelist. At this time, the path planning has been completed. The path is searched reversely from the parent node grid information in the complete information set of all grids in the close list. The reverse connection constitutes the optimized path. The path planning is completed and saved. If the target grid is not added to the close list, the path planning calculation continues.

[0031] Furthermore, in the above step 3.4, according to the grid N i The grid matrix S is mapped with the safe braking distance L to obtain the mapped grid matrix S'. The specific strategy is:

[0032] Move the center grid M of the robot's entity coverage grid matrix S at the current moment to the center grid M of the grid matrix S at the next moment according to the safe braking distance L. j When , the grid set that the grid matrix S passes through during the entire movement process is used as the mapping grid matrix S'. The specific corresponding mapping method is:

[0033]

[0034] Among them, x S and S is the coordinates of each grid in the grid matrix S covered by the robot at the current moment; N1, N2, N3, N4, N5, N6, N7 and N8 are the coordinates of each grid in the grid matrix M j-1The eight-domain grid is numbered in increasing order from top to bottom and from left to right.

[0035] Furthermore, the path cost F of the grid is calculated in the above step 3.1, specifically:

[0036] The distance between grids in the grid map is used as the cost value, and the A-star algorithm is used to calculate the cumulative cost value G in the planned path from the starting grid to the current grid. The cost value of moving the grid up, down, left, and right by a safe braking distance L is 10, and the cost value of moving the grid to the upper left, upper right, lower left, and lower right is 14;

[0037] Then calculate the Euclidean distance from the current grid to the target state as the cost value H, and add the two together to get the path cost F:

[0038] F=G+H.

[0039] The above method of the present invention is further illustrated by non-limiting examples.

[0040] like Figure 2 As shown, in the regional scene schematic diagram of the embodiment, the search area is divided into a 20*20 grid area, the obstacles are 20 randomly distributed 1×1 grid areas (m=n=20, k=1mm), the position coordinates of the starting point are set to (3,3), the coordinates of the target point are set to (17,18), and the grid points adjacent to the grid where the starting point is located are searched, and can be moved along the nearby 8 areas. This implementation column sets the robot carrier to occupy 3×3 grids, and the robot positioning point is the geometric center of the carrier. The cost of translating one grid is 10, and the cost of slanting one grid is 14. Therefore, first put the initial path planning starting point into the close list, and then start moving from the starting point in the 8 areas. After selecting one point in the area, the set S' of all points covered by the robot in the process of moving from the positioning point to the point is judged. When all points in the set S' are passable, the selected points are put into the open list for F value sorting, and the one with the smallest F value is found as the positioning point of the robot at the next moment, and then this node is put into the close list, and the new moving point is judged again. Similarly, the traversed points are put into the close list and not considered. The process is repeated many times until the target point is found. At this time, the optimal planned route can be obtained.

[0041] like Figures 3-1 to 3-18 As shown, it is a schematic diagram of the mapping grid matrix S' obtained by mapping the grid matrix S. Among them, Figure 3-1 It means that when the map is of low resolution, the robot covers only one grid, that is, the size of the grid matrix S is 1×1, the safe braking distance L is 1 at this resolution, and the eight-domain grid of the central grid M is also shown in the figure; Figure 3-2 to Figure 3-9 It means that when the robot moves from the central grid M to the 1st to 8th area grids according to the safe braking distance L, all the grids covered during the movement of S are represented as the mapping grid matrix S'. All the dark grids in the figure represent S'. Figure 3-10 This means that when the map is at high resolution, the robot covers an area of ​​3×3 grids, the size of the grid matrix S is 3×3, the safe braking distance L is 1 at this resolution, and the eight-domain grid of the center grid M is also shown in the figure. Figures 3-11 to 3-18 It represents the mapping diagram of the grid matrix S when the robot has a high-resolution map.

[0042] from Figure 2 and Figure 3-1 to Figure 3-18 It is not difficult to see that the present invention has advantages over the traditional A-star algorithm for planning paths: the present invention provides a mobile robot path planning method based on the spatial constraint A-star algorithm, which can accurately plan the path of the robot from the starting point to the target point, while taking into account the physical size of the robot, avoiding the situation where the planned actual route is too narrow and the machine cannot reach it; at the same time, combined with the robot's safe braking distance, it avoids collisions when the machine travels along the planned path, effectively ensuring the safety of the machine's travel; in addition, the algorithm can also flexibly adapt to the requirements of different map resolutions, and can change the path planning scheme according to actual needs, and can cope with path planning problems in a variety of complex environments.

[0043] The above embodiments are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such special statements and embodiments. Within the knowledge of ordinary technicians in this field, various other specific variations and combinations that do not deviate from the essence of the present invention can also be made based on the technical enlightenment disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A mobile robot path planning method based on spatial constraint A-star algorithm, characterized in that: The following steps are involved: Step 1: Construct a grid map based on the actual work scene area and the required map resolution, and mark the grids occupied by scene obstacles in the map as inaccessible grids; Step 2: Combined with the grid map Map established in step 1, determine the robot's physical coverage grid matrix S and safe braking distance L; Step 3: In the A-star algorithm, two constraints, the robot entity coverage grid matrix S and the safe braking distance L, are added to perform path optimization planning to obtain a safe and effective path for the robot from the starting grid to the target grid; the path optimization planning method described in step 3 is: Step 3.1: In the grid map Map, determine the starting grid and target grid of the robot path planning, and use M to represent the center grid of the grid matrix S where the robot entity covers the jth moment. j-1 To represent, where j is an integer, j = 1 ~ the number of grids in path planning, j-1 represents the number of parent nodes of the central grid, and j-1 = 0 when there is no parent node grid; when initialized, j = 1, the initial central grid M0 coincides with the starting grid, and then the complete information set of the initial central grid M0 is put into the closelist to complete the initialization of path planning. The complete information set of the central grid includes: the central grid M j-1 The coordinates of the grid, the parent node grid M (j-1)-1 The coordinates and path cost F of the central grid that has been planned are stored in the close list. The grids in the close list will no longer participate in the subsequent path planning calculation. The partial information set of the passable grid in the path planning is stored in the open list. The partial information set of the passable grid includes: the coordinates of the grid and the path cost F. Step 3.2: Determine the grid M j-1 Adjacent eight-domain grid N i The ith grid in the grid, i is an integer from 1 to 8, whether it is an inaccessible grid or a grid in the closelist. If it is an inaccessible grid, the grid is ignored. i ; If it is a passable grid, proceed to the next step; Step 3.3: According to the accessible grid N i The grid matrix S is mapped with the safe braking distance L to obtain the mapped grid matrix S', and then it is determined whether there is an impassable grid in the grid matrix S'. If so, the grid N is ignored. i Then return to step 3.

2. If it does not exist, proceed to the next step; Step 3.4: Calculate the traversable grid N i The path cost F is used to determine the grid N i Is it in the openlist? If not, put the grid N i The current partial information set is put into the open list. If the grid N i If it already exists in the open list, the grid N i The current partial information set replaces the existing partial information set; Step 3.5: Determine grid N i Is i less than or equal to 8? If so, return to step 3.

2. After traversing all eight domain grids, sort all grids in the open list according to the F value from small to large, and take the grid with the smallest F value as the central grid M of the robot's entity coverage grid matrix S at the next moment. j , and then the grid M j Remove the open list and then move the current center grid M j-1 Set as the next moment grid M j The parent node grid; finally, the grid M j-1 The complete information set is moved into closelist; Step 3.6: Repeat steps 3.2, 3.3, 3.4 and 3.5 until the target grid is added to the close list. At this time, the path planning is completed. The path is searched reversely from the parent node grid information in the complete information set of all grids in the close list. The reverse connection constitutes the optimized path. The path planning is completed and saved. If the target grid is not added to the close list, the path planning calculation continues; In step 3.4, according to the grid N i The grid matrix S is mapped with the safe braking distance L to obtain the mapped grid matrix S'. The specific strategy is: Move the center grid M of the robot's entity coverage grid matrix S at the current moment to the center grid M of the grid matrix S at the next moment according to the safe braking distance L. j When , the grid set that the grid matrix S passes through during the entire movement process is used as the mapping grid matrix S'. The specific corresponding mapping method is: Among them, x S and S is the coordinates of each grid in the grid matrix S covered by the robot at the current moment; N1, N2, N3, N4, N5, N6, N7 and N8 are the coordinates of each grid in the grid matrix M j-1 The eight-domain grid is numbered in increasing order from top to bottom and from left to right.

2. The mobile robot path planning method based on the spatial constraint A-star algorithm according to claim 1 is characterized in that: In step 1, a grid map Map is constructed according to the actual working scene area and the required map resolution, and the grids occupied by scene obstacles in the Map are marked as inaccessible grids, specifically: According to the actual working scene area and the required map resolution m×n, the grid map Map is constructed using the grid method. The working environment is divided into an m×n grid matrix. The length and width of each grid are k. According to the grid position occupied by the scene obstacles in the grid map Map, the inaccessible grids are marked, and the remaining grids are passable grids.

3. The mobile robot path planning method based on the spatial constraint A-star algorithm according to claim 1 is characterized in that: The step 2 combines the grid map Map established in step 1 to determine the robot's physical coverage grid matrix S and safe braking distance L, specifically: The robot's physical coverage grid matrix S is the set of grids occupied by the robot in the grid map Map at the current moment, and the safe braking distance L is the side length k of the grid in the grid map Map.

4. The mobile robot path planning method based on the spatially constrained A-star algorithm according to claim 1 is characterized in that: The path cost F of the grid is calculated in step 3.1, specifically: The distance between grids in the grid map is used as the cost value, and the A-star algorithm is used to calculate the cumulative cost value G in the planned path from the starting grid to the current grid. The cost value of moving the grid up, down, left, and right by a safe braking distance L is 10, and the cost value of moving the grid to the upper left, upper right, lower left, and lower right is 14; then the Euclidean distance from the current grid to the target state is calculated as the cost value H, and the two are added to get the path cost F: F=G+H.