Path planning method of autonomous mobile robot based on improved A* algorithm

Abstract

The invention provides a path planning method of an autonomous mobile robot based on an improved A * algorithm, which improves the efficiency of a global planning algorithm by adjusting the cost weight of a path from a node of the A * algorithm to a target, and then completes local path planning by using the improved DWA algorithm, thereby improving the real-time performance and flexibility of path planning of the autonomous mobile robot. Through fusion of an A * algorithm and a DWA algorithm, in the aspect of global planning, cost functions are adjusted, redundant nodes are reduced, time is shortened, and search efficiency is improved. In the aspect of local planning, the robot reasonably bypasses the obstacle by depending on the distance from the obstacle, thereby avoiding falling into global optimum and improving the moving safety of the robot.

Description

Technical Field

[0001] The present invention relates to the technical field of path planning, and specifically relates to a path planning method for an autonomous mobile robot based on an improved A* algorithm. Background Art

[0002] In recent years, autonomous mobile robots (AMRs) have been commonly used in the logistics transportation and warehousing management links of intelligent manufacturing workshops to improve the efficiency of internal logistics and material handling tasks, and are one of the key technologies for realizing intelligent manufacturing. The path planning problem is a research hotspot in the field of AMRs, and a path planning algorithm is used to find a collision-free path with obstacles. The path planning problem with a completely known environment is global path planning, and the path planning problem with a partially known environment is local path planning. Commonly used path planning methods include: RRT algorithm, artificial potential field method, D* algorithm, A* algorithm, Dijkstra algorithm, and neural network algorithm, etc. Relevant scholars often use global path planning algorithms to find the optimal path, and rely on local path planning to ensure avoiding obstacles in real time.

[0003] The traditional A* algorithm is widely used in global path planning, but the A* algorithm has a slow search speed and many path inflection points, which does not meet the motion requirements of smooth trajectory and stable speed change of AMRs.

[0004] In local path planning, the combination of the dynamic window method is often adopted, but the traditional DWA algorithm has insufficient foresight and cannot avoid obstacles when encountering a "U"-shaped obstacle, and has poor global optimality.

[0005] For example, in the patent document with the patent application number 202311419045.8 and the publication date of January 26, 2024, a hybrid path planning method based on an improved A* algorithm and a dynamic window method is disclosed, belonging to the technical field of robot path planning. This method combines global prediction and local real-time planning into a hybrid two-layer algorithm, that is, an improved A* algorithm is used to complete global prediction to obtain globally optimal path nodes; then inflection points are selected from these nodes as the guidance for real-time planning, and the real-time planning layer uses an improved dynamic window method to dynamically avoid obstacles in real time, continuously runs each local target point, and finally reaches the target point to complete the task. By using the improved A* algorithm, the shortest path can be found in the search space; at the same time, an optimized dynamic window method is used for local path planning, and the window size can be adjusted according to the changes in the dynamic environment, thereby ensuring the safety of the path and improving the practicality and flexibility of the dynamic obstacle avoidance method.

[0006] In the above literature, the improved A* algorithm increases the observation probability in front of the estimated cost, and the observation probability is related to the area of obstacles within the obstacle area. This makes it necessary to calculate the number of nodes and the differences in the horizontal and vertical directions during the calculation process, resulting in complex calculations. In addition, when the number is small within the rectangular area, the observation probability will be small, so that the function of truly improving the estimated cost cannot be realized, and it is easy to have too high a search frequency, leading to complex calculations. In addition, when using the dynamic window method to achieve obstacle avoidance, it does not consider adjusting the speed evaluation weight according to the distance value, so that when the distance is relatively close, a relatively long driving distance is required to reach a speed of 0, resulting in low obstacle avoidance efficiency. Summary of the Invention

[0007] The purpose of the present invention is to provide a path planning method for an autonomous mobile robot based on an improved A* algorithm. By adjusting the path cost weight from the node to the target of the A* algorithm, the efficiency of the global planning algorithm is improved, and then the improved DWA algorithm is used to complete the local path planning, improving the path planning and obstacle avoidance efficiency of the autonomous mobile robot.

[0008] To achieve the above object, a path planning method for an autonomous mobile robot based on an improved A* algorithm specifically includes the following steps: (1) Generate the initial movement trajectory of the robot according to the current map.

[0009] (2) Use the improved A* algorithm to perform global path planning according to the initial movement trajectory. The improved A* algorithm includes: F(n) = G(n) + H(n) * H(n) * H(n); where H(n) is the estimated cost function, the estimated cost function is set as the distance value between the current node and the target node, F(n) is the cost function from the starting point to node n, and G(n) is the actual cost from the starting point to node n. (3) Move based on the global optimal path.

[0010] (4) Detect whether there are unknown obstacles in the path. If there are obstacles, perform local path adjustment through the evaluation function in the improved DWA algorithm. The evaluation function of the improved DWA algorithm is: ; A function for evaluating the deviation between the target direction and the simulated trajectory end at the current speed; is the distance function between the autonomous mobile robot and the obstacle object on the movement trajectory, is the evaluation function of the current movement speed magnitude of the autonomous mobile robot, is the current speed is the weight coefficient, is the distance between the robot and the obstacle object is the weight coefficient, is the normalization parameter for three sub-functions; x and y are the coordinates of the current point, and are the coordinates where the obstacle is located.

[0011] In the above method, the global path planning is formed by adjusting and improving the A* algorithm. In the improved A* algorithm, the current node and the target node are set as distance values, and three times the distance value is set as the estimated cost function, so as to improve the estimated cost function value and prevent the situation that the distance cost function is too small and the search frequency is too large, resulting in high computational complexity. And the calculation is simple. Then, the improved DWA algorithm is used to complete the local path planning. When there is an obstacle, the local path is adjusted by the improved DWA algorithm and adjusted according to the situation. When the distance between the robot and the obstacle is greater than the radius of the obstacle and less than four times the radius of the obstacle, The value of is x and y are the coordinates of the current point, is the coordinate where the obstacle is located. At this time, as the distance decreases, the mobile robot needs to gradually increase the speed variable, so that the difference between the current speed and the target speed increases. The current speed will not directly become the target speed of 0, so a greater running distance is required to directly reduce the speed to 0 in a short time, so as to achieve gradual adjustment to avoid obstacles; when the distance between the robot and the obstacle is greater than four times the radius of the obstacle, The value of is 3. The current value becomes smaller compared to the distance between the current point and the target point, indicating that the mobile robot maintains a safe distance from the obstacle and can continue the current behavior; The above

[0012] Furthermore, step (2) further includes: (21) setting a starting point, an ending point, and more than one node formed between the starting point and the ending point according to the simulated initial movement path of the robot; adding the starting point, the nodes, and the ending point to the openlist list; (22) Use the updated starting point as the parent node and search for nodes in the preset directions around the parent node; (23) Determine whether there are obstacles between the parent node and the nodes around the parent node. If there are no obstacles, proceed to step (24). If there are obstacles, add the current node to the closelist and perform step (21) to reset the starting point, nodes, and end point; (24) Select the node that has no obstacles between it and the parent node and has the shortest straight-line distance from the parent node through the improved A* algorithm, and form a path between the parent node and this node; (25) Update the node in step (24) as the starting point; (26) Search whether there is an end point in the directions around the updated starting point; (27) If there is no end point, repeat steps (22)-(26).

[0013] (28) If there is an end point, connect the updated starting point and the end point to form the globally optimal path.

[0014] The above settings search for nodes in multiple directions around the parent node and, when encountering obstacles, evaluate the nodes closer to the parent node according to the improved A* algorithm as child nodes to form the global path. The operation method is simple and reliable.

[0015] Further, after step (4), it also includes (5) After avoiding unknown obstacles, repeat steps (2)-(4) to move to the end point.

[0016] The above settings repeat the operation after avoiding unknown obstacles to move to the end point, ensuring accurate movement to the end point.

[0017] Further, in step (4), the estimated cost function H(n) is calculated as follows: .

[0018] In the formula, is the target node coordinate, is the current node coordinate.

[0019] The above settings calculate the estimated cost function H(n) to ensure that the node with the shortest straight-line distance to the target point can be preferentially selected, improving the efficiency and flexibility of the algorithm.

[0020] Further, step (4) specifically includes: (41) Use the improved DWA algorithm to perform spatial sampling on the robot's speed.

[0021] (42) Constrain the robot's speed.

[0022] (43) Constrain the acceleration and deceleration of the robot motor.

[0023] (44) Constrain the braking distance of the robot.

[0024] (45) Simulate the motion trajectory of the robot and use the evaluation function to calculate the optimal trajectory.

[0025] With the above settings, through the improved DWA algorithm, the robot is optimized respectively in terms of speed, acceleration and deceleration, and braking distance, so as to save the obstacle avoidance time of the robot, improve the obstacle avoidance efficiency, reduce the number of trajectory turning points, and make the operation of the robot more stable.

[0026] Further, the speed constraint formula in step (42) is: .

[0027] In the formula, v is the linear velocity of the robot, w is the angular velocity of the robot, are the maximum and minimum values of the linear velocity, are the maximum and minimum values of the angular velocity.

[0028] With the above settings, by constraining the speed of the robot, it is convenient to simulate the motion trajectory of the robot.

[0029] Further, the maximum and minimum acceleration sets for acceleration and deceleration constraints in step (43) are:

[0030] In the formula, are the current velocity and angular velocity, are the maximum acceleration and angular acceleration, are the maximum deceleration and deceleration angular velocity of the robot.

[0031] With the above settings, by constraining the acceleration and deceleration of the robot, it is convenient to simulate the motion trajectory of the robot.

[0032] Further, the braking distance constraint in step (44) is:

[0033] In the formula, the function item dist(v, w) is used to evaluate the distance between the trajectory corresponding to the current velocity of the robot and the nearest obstacle.

[0034] With the above settings, by constraining the braking distance of the robot, the robot needs to decelerate in time when encountering an obstacle during movement and stop immediately before hitting the obstacle, preventing safety accidents.

[0035] Further, the improved A* algorithm in step (2) includes: when H(n) is less than a preset value, F(n) = G(n) + H(n) * H(n) * H(n); when H(n) is greater than the preset value, F(n) = G(n) + H(n) * H(n).

[0036] With the above settings, when the distance value is small, the cube of H(n) is used to increase it, and when it is less than the preset value, twice H(n) is used, thus preventing the situation where the weight coefficient increases too much, resulting in an inaccurate search range due to too small a search scope. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a schematic diagram for comparing the path planning of the improved A* algorithm and the traditional A* algorithm of the present invention.

[0038] Figure 2 It is a schematic diagram for comparing the planning of the improved DWA algorithm and the traditional DWA algorithm of the present invention.

[0039] Figure 3 It is a schematic diagram of the trajectory of the mobile robot planned by the fusion algorithm of the present invention.

[0040] Figure 4 It is a schematic diagram of the changes in the speed and angular velocity of the mobile robot in the simulation experiment of the present invention.

[0041] Figure 5 It is a test schematic diagram of the fusion algorithm of the present invention in a static obstacle scenario.

[0042] Figure 6 It is a schematic diagram of the changes in the speed and angular velocity of the mobile robot of the present invention in a static obstacle scenario.

[0043] Figure 7 It is a test schematic diagram of the fusion algorithm of the present invention in a dynamic obstacle scenario.

[0044] Figure 8 It is a schematic diagram of the changes in the speed and angular velocity of the mobile robot of the present invention in a dynamic obstacle scenario.

[0045] Figure 9 It is a schematic diagram of the starting point repeated positioning accuracy of the mobile robot of the present invention.

[0046] Figure 10 It is a schematic diagram of the end point repeated positioning accuracy of the mobile robot of the present invention.

[0047] Figure 11 It is a flowchart of the operation of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0048] The present invention will be further described in detail below in conjunction with the drawings and specific embodiments.

[0049] As Figures 1-11 shown, the present invention provides a path planning method for an autonomous mobile robot based on an improved A* algorithm. In this embodiment, this method is applied to a mobile robot for an intelligent photovoltaic cell production workshop. The mobile robot uses a XingSong LE-50621-V1 lidar to scan and obtain map information, and uses a Demark DMLHO13-A5 motor to provide power. Considering the comprehensive application scenario requirements, the parameters of the mobile robot are shown in Table 1.

[0050]

[0051] Table 1 The motion parameters of the mobile robot are shown in Table 2:

[0052] Table 2 The mobile robot adopts a two-wheel differential model, consisting of two driving wheels and four driven wheels; the kinematic model of the mobile robot is: .

[0053] In the formula, is the deflection angle of the robot relative to the global coordinate system, is the coordinate position of the mobile robot at time k +1, are the linear and angular position increments of the mobile robot at time k +1, respectively.

[0054] As Figures 1-11 shown, a path planning method for an autonomous mobile robot based on an improved A* algorithm specifically includes the following steps: (1) Generate an initial movement trajectory of the robot according to the current map.

[0055] (2) Use the improved A* algorithm to perform path planning according to the movement track.

[0056] (21) Set a starting point, an ending point, and more than one node formed between the starting point and the ending point according to the simulated movement trajectory of the robot; add the starting point, the node, and the ending point to the openlist list.

[0057] (22) Update the starting point as the parent node and search for the nodes in the surrounding directions of the parent node.

[0058] (23) Determine whether there is an obstacle between the parent node and the nodes around the parent node. If there is no obstacle, perform step (24). If there is an obstacle, add the current node to the closelist list and perform step (21) to reset the starting point, the node, and the ending point.

[0059] (24) Perform global path planning according to the improved A* algorithm based on the initial movement trajectory. The improved A* algorithm includes: F(n) = G(n) + H(n) * H(n) * H(n); where H(n) is the estimated cost function, and the estimated cost function is set as the distance value between the current node and the target node. F(n) is the cost function from the starting point to node n, and G(n) is the actual cost from the starting point to node n. Specifically, select the node that has no obstacle between it and the parent node and has the shortest straight-line distance, and form a path between the parent node and this node.

[0060] (25) Update the node in step (24) as the starting point.

[0061] (26) Search whether there is an end point in the directions around the updated starting point.

[0062] (27) If there is no end point, repeat steps (22)-(26).

[0063] (28) If there is an end point, connect the updated starting point and the end point to form the globally optimal path.

[0064] (3) The robot moves based on the globally optimal path.

[0065] (4) Detect whether there are unknown obstacles on the path. If there are obstacles, perform local path adjustment through the evaluation function in the improved DWA algorithm. The evaluation function of the improved DWA algorithm is: ; ; A function used to evaluate the deviation between the target direction and the end of the simulated trajectory at the current speed; is the distance function between the mobile robot and the obstacle object on the motion trajectory, is the evaluation function of the current motion speed magnitude of the mobile robot, is the current speed is the weight coefficient, is the distance between the robot and the obstacle object is the weight coefficient, is the normalization parameter of the three sub-functions; x and y are the coordinates of the current point, and are the coordinates where the obstacle is located.

[0066] Step (4) specifically includes: (41) Use the improved DWA algorithm to perform spatial sampling on the speed of the robot.

[0067] (42) Constrain the speed of the robot.

[0068] (43) Constrain the acceleration and deceleration of the robot's motor.

[0069] (44)Constrain the braking distance of the robot.

[0070] (45)Simulate the motion trajectory of the robot and use the evaluation function to calculate the optimal trajectory.

[0071] (5)After avoiding unknown obstacles, repeat steps (2)-(4) to move to the end point.

[0072] In this embodiment, the improved A* algorithm in step (2) is as follows: F(n) = G(n) + H(n) * H(n) * H(n) (1).

[0073] In the photovoltaic cell production scenario, the space is large and the obstacles are complex. The traditional A* algorithm has strong symmetry and traverses many nodes, resulting in too long calculation time, getting stuck in local optimality, and reducing the algorithm efficiency. This embodiment proposes to increase the weight of H(n) and preferentially select the node with the shortest straight-line distance to the target point to improve the efficiency and flexibility of the algorithm.

[0074] Use the Euclidean distance calculation formula to calculate the estimated cost function H(n): (2).

[0075] In the formula, is the coordinate of the target node, is the coordinate of the current node.

[0076] Use the MATLAB software to set grid maps of different range sizes to simulate the A* algorithm, and record the path planning process of the A* algorithm before and after improvement. As Figure 1 shown, (a) is the traditional A* algorithm planning in 30*30m; (b) is the improved A* algorithm planning in 30*30m; (c) is the traditional A* algorithm planning in 40*40m; (d) is the improved A* algorithm planning in 40*40m.

[0077] Record the number of calculation steps, distance, time, and the number of redundant nodes of the A* algorithm before and after improvement in the path. The results are shown in Table 3 below:

[0078] Table 3 Combined with the charts, it can be seen that compared with the traditional A* algorithm, in the 30*30m and 40*40m maps, the running distances of the improved A* algorithm are shortened by 2.68% and 0.72% respectively, the running times are shortened by 454% and 371.2% respectively, and the number of nodes is greatly reduced. In summary, compared with the traditional A* algorithm, the improved A* algorithm takes less time, reduces the search nodes, and improves the path planning efficiency.

[0079] In another embodiment, the improved A* algorithm in step (2) includes: when H(n) is less than a preset value, F(n) = G(n) + H(n) * H(n) * H(n); when H(n) is greater than the preset value, F(n) = G(n) + H(n) * H(n). For example, when the preset value is 10m, the cube of H(n) is used to increase it when the distance value is small, and when it is less than the preset value, twice H(n) is used, so as to prevent the situation that the search range becomes too small due to the excessive increase of the weight coefficient, resulting in inaccuracy.

[0080] In this embodiment, the DWA algorithm is a commonly used local obstacle avoidance planning method. When a mobile robot navigates under a global map with complete environmental information, it will be affected by suddenly emerging dynamic obstacles. The above A* algorithm cannot perform real-time local path planning. Therefore, the real-time obstacle avoidance algorithm DWA needs to be used to quickly plan a local obstacle avoidance path to avoid obstacles that may collide with the mobile robot in the speed search space, so that the mobile robot has the ability of real-time obstacle avoidance.

[0081] DWA samples the speed space of the mobile robot and simulates the motion trajectory of the mobile robot. Affected by the parameters of the physical robot, motor performance, and braking distance, the sampling speed of the robot will be limited within a certain range; The speed constraint formula in step (42) is: (3).

[0082] In the formula, v is the linear velocity of the robot, w is the angular velocity of the robot, are the maximum and minimum values of the linear velocity, are the maximum and minimum values of the angular velocity.

[0083] The maximum and minimum acceleration sets of the acceleration and deceleration constraints in step (43) are: (4).

[0084] In the formula, are the current velocity and angular velocity, are the maximum acceleration and angular acceleration, are the maximum deceleration and deceleration angular velocity of the robot.

[0085] The braking distance constraint in step (44) is: (5).

[0086] In the formula, the function item dist(v, w) is used to evaluate the distance between the trajectory corresponding to the current velocity of the robot and the nearest obstacle.

[0087] By restricting the braking distance of the robot, the robot can decelerate in time when encountering an obstacle during movement and stop immediately before hitting the obstacle, preventing safety accidents.

[0088] In the traditional DWA algorithm, the evaluation function is: (6).

[0089] Among them, the function item head(v, w) represents the consistency between the heading of the robot at the end of the trajectory corresponding to the current speed and the target direction. is a smoothing function, and the function item vel(v, w) is used to evaluate the motion performance of the robot. The higher the speed, the higher the score. are the weight coefficients of each function item.

[0090] The traditional DWA algorithm needs to evaluate a large number of optional trajectories, and the evaluation weight parameters are fixed, so it cannot flexibly handle obstacles that may appear at any time. For example, when an obstacle suddenly appears near the target point, when the mobile robot is about to reach the target point, there will be a situation where the movement is very slow or even stuck. In addition, in the DWA algorithm, the weight of the function item is always less than the weights of other evaluation sub-functions, that is, the speed in the algorithm cannot guarantee to reach the end point quickly and smoothly, so that the trajectory cannot meet the optimal efficiency.

[0091] Therefore, by relying on the distance from the obstacle, the size of the speed evaluation weight is dynamically changed. The farther away from the obstacle, the smaller the weight value and the smoother the speed. The closer to the obstacle, it can quickly avoid the obstacle. The improved evaluation function is: (7).

[0092] Among them, .

[0093] ; Set the dynamic weight value according to the distance from the obstacle as .

[0094] Using MATLAB software for simulation, in the simulation state, set the obstacle radius R = 0.5m, set a = 0.4, b = 0.1, so that ranges from 0.1 to 0.4. Based on MATLAB, compare the simulation of the DWA algorithm before and after improvement in the same map environment. The starting point is (0, 0) as Figure 2 shown, where (a) is the 30*30m traditional DWA algorithm planning; (b) is the 30*30m improved DWA algorithm planning; (c) is the 40*40m traditional DWA algorithm planning; (d) is the 40*40m improved DWA algorithm planning. The comparison between the improved DWA algorithm and the traditional DWA algorithm is shown in Table 4 below:

[0095] Table 4 Combined with Figure 2 and Table 4, it can be obtained that for local path planning, compared with the traditional DWA algorithm, the improved DWA algorithm shortens the running distance by 23.91% and 4.71% respectively in the 30*30m and 40*40m maps, reduces the number of path turns by 33% for both, and reduces the running time by 22.04% and 14.66% respectively. In summary, the improved DWA algorithm saves time, improves the obstacle avoidance efficiency, reduces the number of trajectory turns, and makes the robot move more stably.

[0096] Through MATLAB software, a fusion algorithm simulation is carried out on the improved A* algorithm and the improved DWA algorithm. In global planning, the cost function is adjusted to reduce redundant nodes, shorten the time, and improve the search efficiency. In local planning, it reasonably bypasses obstacles depending on the distance from the obstacles to avoid falling into the global optimum and improve the moving safety of the robot. In the established simulation environment, first determine the positions of the target point and the starting point, and apply the fusion algorithm for testing and verification. Record the trajectory planning of the fusion algorithm when the mobile robot faces static obstacles and dynamic obstacles respectively.

[0097] The trajectory of the mobile robot planned by the fusion algorithm is as Figure 3 shown. In the figure, the dark squares are static obstacles, the light squares are dynamic obstacles, the semi-transparent squares are the end points, and the dotted lines are the trajectories planned by the initial fusion algorithm. When a dynamic obstacle appears, the fusion algorithm re-plans the path to bypass the obstacle, and then continues to move forward along the globally optimal path planned initially.

[0098] Record the changes in the speed and angular velocity of the mobile robot in the simulation experiment as Figure 4 shown. It can be seen from the figure that the linear velocity of the robot increases steadily and then remains stable. When encountering an obstacle, the speed decreases and then increases steadily. It can be seen that the robot moves stably and the speed does not change suddenly. The angular velocity fluctuates greatly, indicating that the robot adjusts its direction in real time during the obstacle avoidance process and has a strong obstacle avoidance ability.

[0099] The fusion algorithm is further tested on a prototype in a real environment. In the static obstacle scenario, an integrated mobile robot is used to verify the fusion algorithm, and the process of detecting obstacles and planning obstacle avoidance trajectories in the ROS system as well as the movement of the mobile robot in the real scenario are recorded. The test site is about 64 square meters. Five obstacles of different sizes are set, represented by serial numbers 1 to 5. As Figure 5As shown, the mobile robot starts from the starting point 01, detects obstacles 1 and 2, plans a path, turns left to bypass obstacles 1 and 2; detects obstacle 3, plans a path, turns right to bypass obstacle 3; detects obstacles 4 and 5, plans a path, turns left to bypass obstacles 4 and 5, and then reaches the end point 02.

[0100] As Figure 6 shown; record the changes in the speed and angular velocity of the mobile robot. During the obstacle avoidance movement, the linear velocity of the mobile robot rises smoothly until it encounters an obstacle, and then the linear velocity decreases smoothly and is continuously adjusted. The angular velocity is in a fluctuating change, proving that the algorithm reduces the turning angle when encountering an obstacle.

[0101] As Figure 7 shown; in the dynamic obstacle scenario, five static obstacles with different sizes and an experimenter as a dynamic obstacle are also set. The experiments are represented by serial numbers 1 to 6, where serial number 3 is the dynamic obstacle. Record the movement process of the mobile robot. While ensuring the safety of the experimenter, the mobile robot starts from the starting point, detects obstacles 1 and 2, plans a path, turns left to bypass obstacles 1 and 2; detects the dynamic obstacle 3 and the static obstacle 4 that appear ahead, retreats and replans the path, turns right to bypass obstacles 3 and 4; detects obstacles 5 and 6, plans a path, turns left to bypass obstacles 5 and 6, and then reaches the end point.

[0102] As Figure 8 shown, record the speed and angular velocity of the robot. The linear velocity of the mobile robot rises smoothly until it encounters an obstacle, and then the linear velocity decreases smoothly and is continuously adjusted. The change in angular velocity is significantly fluctuating, proving that the fusion algorithm adjusts the obstacle avoidance route in real time to ensure the path safety of the mobile robot.

[0103] In the photovoltaic cell production workshop, the robot needs to transport materials back and forth multiple times, and accurate positioning can ensure the safety of material docking. Design a repeated positioning experiment to record the repositioning accuracy of the fusion algorithm. Record the reciprocating movement of the mobile robot between the starting point A and the end point B, repeat the experiment 16 times, and use the coordinates of the mobile robot's first arrival at points A and B recorded as reference values to calculate the repeated positioning accuracy of the mobile robot at points A and B. As Figure 9 and Figure 10 shown, Figure 9 is the repeated positioning accuracy at point A, Figure 10 is the repeated positioning accuracy at point B.

[0104] It can be seen that the average error of the mobile robot in the X direction at points A and B is 20 mm, and the average error in the Y direction at points A and B is 15 mm. This proves that the movement of the mobile robot planned by the fusion algorithm has good positioning accuracy.

[0105] The above prototype experiments and repeated positioning experiment results show that in a real and complex environment, the integration of the improved A* algorithm and the improved DWA algorithm enables the robot to have the ability of global path planning and dynamic obstacle avoidance in static and dynamic obstacle scenarios, and has a high precision in motion repositioning, meeting the material transportation requirements of the photovoltaic cell production workshop and ensuring the safety of the mobile robot itself, the staff, and the goods.

[0106] The evaluation function of the traditional A* algorithm is improved by adjusting the path cost weight from the node to the target. The simulation results show that in the 30*30m and 40*40m maps, the running distances of the improved A* algorithm are shortened by 2.68% and 0.72% respectively, the running times are shortened by 454% and 371.2% respectively, and the number of nodes is significantly reduced. The DWA algorithm is improved by changing the speed evaluation weight to complete the real-time dynamic obstacle avoidance of the mobile robot. The simulation results show that in the 30*30m and 40*40m maps, the running distances of the improved DWA algorithm are shortened by 23.91% and 4.71% respectively, the number of path turning times is reduced by 33% for both, and the running times are reduced by 22.04% and 14.66% respectively. The fusion algorithm of the improved A* algorithm and the improved DWA algorithm enables the mobile robot to have the ability of global path planning and dynamic obstacle avoidance in a complex environment. In the simulation environment test, the mobile robot can quickly navigate and plan to reach the target point. When the robot encounters unknown obstacles and pedestrians, it can re-plan the route to avoid the obstacles and return to the global path, achieving the function of real-time obstacle avoidance. At the same time, multiple repositioning experiments show that the average error of the robot in the longitudinal direction (X-axis direction) is ≤20mm, and the average error in the transverse direction (Y-axis) is ≤15mm, indicating that the fusion algorithm has a high navigation accuracy and meets the material handling requirements of the photovoltaic panels.

[0107] The beneficial effects of the present invention: By adjusting the path cost weight from the A* algorithm node to the target, the efficiency of the global planning algorithm is improved. Then, the improved DWA algorithm is used to complete the local path planning, improving the real-time performance and flexibility of the path planning of the autonomous mobile robot. Through the integration of the A* algorithm and the DWA algorithm, in terms of global planning, the cost function is adjusted to reduce redundant nodes, shorten the time, and improve the search efficiency. In terms of local planning, depending on the distance from the obstacle, the obstacle is reasonably bypassed to avoid falling into the global optimum and improve the moving safety of the robot.

Claims

1. A path planning method for an autonomous mobile robot based on an improved A* algorithm, characterized in that: The specific steps include: (1) Generate the robot’s initial movement trajectory based on the current map; (2) Use the improved A* algorithm to perform global path planning based on the initial moving trajectory. The improved A* algorithm includes: F(n)=G(n)+H(n)*H(n)*H(n); where H(n) is the estimated cost function, the estimated cost function is set to the distance between the current node and the target node, F(n) is the cost function from the starting point to node n, and G(n) is the actual cost from the starting point to node n; (3) Move based on the global optimal path; (4) Check whether there are unknown obstacles on the path. If there are obstacles, adjust the local path by improving the evaluation function in the DWA algorithm. The evaluation function of the improved DWA algorithm is: ; A function for evaluating the deviation of the target direction from the simulated trajectory end at the current speed; is the distance function between the mobile robot and the obstacle object on the motion trajectory, is the evaluation function of the current movement speed of the mobile robot, Current speed The weight coefficient of is the distance between the robot and the obstacle The weight coefficient of are the normalized parameters of the three sub-functions; x and y are the coordinates of the current point, and are the coordinates of the obstacle.

2. The path planning method for an autonomous mobile robot based on an improved A* algorithm according to claim 1, characterized in that: Step (2) further includes: (21) setting a starting point, an end point, and one or more nodes formed between the starting point and the end point according to the simulated initial moving path of the robot; adding the starting point, the node, and the end point to an openlist list; (22) Update the starting point as the parent node and search for nodes in the preset direction around the parent node; (23) Determine whether there is an obstacle between the parent node and the nodes around the parent node. If there is no obstacle, proceed to step (24). If there is an obstacle, add the current node to the closelist list and proceed to step (21) to reset the starting point, node, and end point. (24) Select a node that has no obstacles between it and the parent node and has the shortest straight-line distance to the parent node through the improved A* algorithm, and form a path between the parent node and the node; (25) Update the node in step (24) to the starting point; (26) Search whether there is an end point in the direction around the updated starting point; (27) If there is no endpoint, repeat steps (22)-(26); (28) If there is an end point, connect the updated start and end points to form a global optimal path.

3. The path planning method for an autonomous mobile robot based on an improved A* algorithm according to claim 1, characterized in that: After step (4), the method further includes (5) avoiding unknown obstacles and repeating steps (2) to (4) to move to the end point.

4. The path planning method for an autonomous mobile robot based on an improved A* algorithm according to claim 3, characterized in that: In step (4), the estimated cost function H(n) is calculated: ; In the formula, is the target node coordinate, is the current node coordinate.

5. The path planning method for an autonomous mobile robot based on an improved A* algorithm according to claim 1, characterized in that: Step (4) specifically includes: (41) Use the improved DWA algorithm to perform spatial sampling of the robot's velocity; (42) Constrain the speed of the robot; (43) Constrain the acceleration and deceleration of the robot motor; (44) Constrain the braking distance of the robot; (45) Simulate the robot's motion trajectory and use the evaluation function to calculate the optimal trajectory.

6. The path planning method for an autonomous mobile robot based on an improved A* algorithm according to claim 5, characterized in that: The speed constraint formula in step (42) is: Where v is the linear velocity of the robot, w is the angular velocity of the robot, are the maximum and minimum values ​​of the linear velocity, are the maximum and minimum angular velocity.

7. The path planning method for an autonomous mobile robot based on an improved A* algorithm according to claim 5, characterized in that: The maximum and minimum acceleration sets of the acceleration and deceleration constraints in step (43) are: ； In the formula, is the current speed and angular velocity, are the maximum acceleration and angular acceleration, are the robot's maximum deceleration and angular velocity.

8. The path planning method for an autonomous mobile robot based on an improved A* algorithm according to claim 5, characterized in that: The braking distance constraint in step (44) is: ； Wherein, the function term dist(v, w) is used to evaluate the distance between the trajectory corresponding to the current speed of the robot and the nearest obstacle.

9. The path planning method for an autonomous mobile robot based on an improved A* algorithm according to claim 1, characterized in that: The improved A* algorithm in step (2) includes: when H(n) is less than a preset value, F(n)=G(n)+H(n)*H(n)*H(n); when H(n) is greater than a preset value, F(n)=G(n)+H(n)*H(n).

Images