Mobile robot path planning method based on improved adaptive dynamic window method

Abstract

The invention provides a mobile robot path planning method based on an improved adaptive dynamic window method, and relates to the technical field of mobile robot path planning. Firstly, a high-precision grid map of a robot working environment is established, an improved A-star algorithm is utilized to generate a global optimal path, redundancy elimination optimization and arc interpolation smoothing processing are carried out, and then an optimal path is selected according to an evaluation function of an improved adaptive dynamic window method and a deflection degree screening mechanism. The improved A-star algorithm and the improved dynamic window method are fused to obtain a fusion algorithm having the advantages of the two algorithms, the improved A-star algorithm improves the algorithm search efficiency, and redundancy elimination optimization and arc interpolation smoothing improve the smoothness of the path. And by using the self-adaptive dynamic window method based on deflection degree screening, the speed of the algorithm is improved, obstacles can be avoided more effectively, the running speed of the robot in a narrow channel is improved, and the running time of path planning is effectively shortened.

Description

Technical Field

[0001] The present invention relates to the technical field of mobile robot path planning, and in particular to a mobile robot path planning method based on an improved adaptive dynamic window method. Background Art

[0002] As an important part of robotics technology, the research and development of mobile robots are of great significance in many fields. In the past few decades, mobile robots have been successfully applied in different fields such as military, industry and security environment. The core function of mobile robots is the path planning of robots. Path planning allows robots to plan a collision-free and effective road from the starting point to the end point in a complex environment. The purpose of path planning algorithm optimization is to improve the path safety, the movement efficiency and stability of the robot posture under the premise of completing path planning. The current mainstream local path planning algorithms are: dynamic window method, gradient method, artificial potential field method, etc. However, most algorithms do not consider the kinematic performance constraints of the mobile robot. The gradient method plans the path according to the direction of gradient descent, and does not take into account the actual movement speed of the mobile robot, the steering performance limitations of the mobile robot, etc. Using the gradient method for path planning in a complex environment is likely to generate a path that the mobile robot may not be able to travel; the artificial potential field law treats the mobile robot as a particle for path planning; the dynamic window law considers the actual kinematic performance control of the mobile robot, and the generated path is more feasible in theory, while the parameters of the evaluation function of the traditional dynamic window method are fixed, and using the same evaluation function parameters in different environments may not result in a satisfactory optimal solution. Summary of the invention

[0003] Aiming at the shortcomings of the traditional dynamic window method, which has the disadvantages of poor environmental adaptability and slow mobilization speed of mobile robots due to the fixed evaluation function parameters, a mobile robot path planning method based on the improved adaptive dynamic window method is proposed, which effectively reduces the path planning time and improves the path planning efficiency. The mobile robot path planning method based on the improved adaptive dynamic window method is combined with the improved A-star algorithm into a fusion algorithm for path planning, so that the planned path is close to the optimal one.

[0004] A mobile robot path planning method based on an improved adaptive dynamic window method comprises the following steps:

[0005] Step 1: Collect the robot's working environment information and establish a working environment map of the mobile robot;

[0006] Step 2: Set the starting point and target point of the mobile robot;

[0007] Step 3: Divide the working environment map of the mobile robot into grids, and determine the starting point grid and the target point grid;

[0008] Step 4: Take the starting point as the parent node, expand the domain child nodes of the parent node and put them into the list of candidate path nodes;

[0009] Step 5: Use the improved cost function in the A-star algorithm to calculate the cost of the domain sub-nodes in the candidate path node list, select the sub-node with the minimum cost according to the cost calculation and move it into the path node list of the A-star algorithm;

[0010] The cost function of the A-star algorithm is improved, and the improved cost function in the A-star algorithm is used for cost calculation. The improved cost function is shown in the following formula:

[0011] f(n)=g(n)+w(n)·h(n)

[0012] Where n is the current node position; f(n) is the cost function; g(n) is the actual cost of reaching the current node n from the starting point; h(n) is the heuristic function; w(n) is the weight function of the heuristic function h(n);

[0013] The calculation of the weight function is shown in the following formula:

[0014]

[0015] Among them, D star (n) is the distance from the starting point to the current node n, D goal (n) is the distance from the current node n to the target point, α and β are parameters that can be adjusted according to the actual planning area;

[0016] Step 6: Determine whether the current node has reached the target point. If not, take the current node as the parent node and expand the domain child nodes of the parent node, put them into the list of candidate path nodes, and repeat steps 5 to 6 until the target point is reached;

[0017] Step 7: Reversely traverse the nodes in the path node list of the A-star algorithm to obtain a global optimal path from the starting point to the target point;

[0018] Step 8: Optimize the global optimal path by removing redundant nodes to obtain an optimized path without redundant nodes;

[0019] Step 8.1: Group the nodes in the global optimal path and determine whether each group of nodes is on the same straight line. If so, skip the middle node in each group of nodes to simplify the path.

[0020] Step 8.2: If each group of nodes is not on the same straight line, calculate the equation of the straight line between the nodes to determine whether there is an obstacle on the straight line. If there is no obstacle, directly connect the first node and the last node to simplify the path. Repeat step 8.1 to continuously detect nodes that meet the conditions for optimization;

[0021] Step 9: Perform arc interpolation processing on the turning corners in the optimized path obtained in step 8 to generate an optimized path;

[0022] Step 9.1: Loop through the coordinates of the nodes in the optimized path, perform arc interpolation processing on the turning path in the optimized path, and obtain an arc path;

[0023] Step 9.2: Check whether the obstacle exists within the arc range. If the generated arc path collides with the obstacle, adjust the end point position of the line segment to generate a new arc path;

[0024] Step 10: Segment the new arc path after the arc interpolation processing in step 9 to generate local target points;

[0025] Step 11: Use the local target points generated in step 10 as the multi-segment local targets of the improved adaptive dynamic window method, generate a speed combination based on the target point position, perform speed sampling, and generate multiple simulated motion trajectories for the next time node;

[0026] Step 12: Select the optimal path according to the evaluation function and deflection screening mechanism of the improved adaptive dynamic window method;

[0027] Step 12.1: By improving the weight of the speed evaluation function, the evaluation function in the adaptive dynamic window method is improved, and the path of the mobile robot is dynamically adjusted using the improved evaluation function. The improved evaluation function is shown in the following formula:

[0028] G(v,ω)=σ(α*heading(v,ω)+β*dist(v,ω)+γd*vel(v,ω))

[0029] Among them, G(v,ω) is the improved evaluation function, σ is the normalization processing, α is the weight of the heading angle evaluation function, heading(v,ω) is the heading angle evaluation function, β is the weight of the obstacle distance evaluation function, dist(v,ω) is the obstacle distance evaluation function, γd is the weight of the improved speed evaluation function, vel(v,ω) is the speed evaluation function, v is the linear velocity of the mobile robot, and ω is the angular velocity of the mobile robot;

[0030] The weight calculation of the improved speed evaluation function is shown in the following formula:

[0031]

[0032]

[0033] Where Dist is the minimum obstacle distance; ε and is an adjustable parameter; x d is the horizontal coordinate of the nearest obstacle, x n is the horizontal coordinate of the mobile robot, y d is the ordinate of the nearest obstacle, y n is the vertical coordinate of the mobile robot; R is the radius of the obstacle;

[0034] Step 12.2: Use the dynamic weight judgment mechanism of deflection degree screening to assist in adjusting the weight of the speed evaluation function;

[0035] In the dynamic adjustment of the path, a dynamic weight judgment mechanism for deflection degree screening is added to calculate the deviation of the orientation angle between the nearest obstacle and the mobile robot; the orientation angle of the mobile robot is calculated and converted into an angle value, the azimuth angle from the mobile robot to the nearest obstacle is calculated and converted into an angle value, and the two angle values ​​obtained above are compared to obtain the direction difference between the robot and the nearest obstacle; if the direction difference exceeds 90°, the mobile robot resumes the high-speed motion state and does not continue to dynamically adjust the path of the mobile robot according to the nearest obstacle in step 12.1. When the mobile robot approaches the next nearest obstacle with a deflection degree difference of less than 90°, step 12.1 is performed again, and the path of the mobile robot is dynamically adjusted again;

[0036] Step 12.3: Select the optimal path according to the evaluation function and deflection screening mechanism of the improved adaptive dynamic window method;

[0037] Step 12.4: The mobile robot moves according to the optimal path, and determines whether the mobile robot reaches the target point. If the mobile robot does not reach the target point, steps 11 to 12 are repeated until the robot reaches the target point according to the optimal path.

[0038] Step 13: Get the complete optimized path planning and simulated motion line.

[0039] The beneficial effects of adopting the above technical solution are:

[0040] 1) The heuristic function of the A-star algorithm is improved, which improves the search efficiency of the A-star algorithm;

[0041] 2) The path planned by the improved A-star algorithm is de-redundanted, and arc interpolation is used on the global path to improve the smoothness of the path;

[0042] 3) The fixed speed weight parameter of the traditional adaptive dynamic window method (DWA) is improved to a dynamic weight parameter related to the obstacle distance, which improves the obstacle avoidance sensitivity of the mobile robot;

[0043] 4) The judgment mechanism based on deflection screening is used to assist in adjusting the speed weight, which improves the operating efficiency of the overall algorithm, increases the operating speed of the mobile robot, and effectively reduces the overall planning time.

[0044] The use of the improved adaptive dynamic window method can effectively improve the speed of the mobile robot and avoid obstacles more effectively in complex terrain environments. The combination of the improved A-star algorithm and the improved adaptive dynamic window method can effectively give play to the advantages of the A-star algorithm and the traditional adaptive dynamic window method (DWA), and can effectively avoid the shortcomings of the A-star algorithm in dynamic obstacle avoidance and the shortcomings of the traditional adaptive dynamic window method (DWA) that it is easy to fall into local optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 A flow chart of a mobile robot path planning method using an improved adaptive dynamic window method provided by an embodiment of the present invention;

[0046] Figure 2 A schematic diagram comparing the improved A-star algorithm provided by an embodiment of the present invention and the original A-star global path;

[0047] Figure 3 A sample diagram of the test environment for the improved adaptive dynamic window method provided by an embodiment of the present invention;

[0048] Figure 4 The improved adaptive dynamic window method provided by the embodiment of the present invention is different from the traditional adaptive dynamic window method (DWA) in that Figure 3 Experimental results under the same experimental environment, where (a) is a comparison of the linear velocity and time of the traditional adaptive dynamic window method (DWA), (b) is a comparison of the angular velocity and time of the traditional adaptive dynamic window method (DWA), (c) is a comparison of the linear velocity and time of the improved adaptive dynamic window method, and (d) is a comparison of the angular velocity and time of the improved adaptive dynamic window method;

[0049] Figure 5 A test result diagram of the fusion algorithm provided in an embodiment of the present invention;

[0050] Figure 6 This is a graph of the linear velocity and angular velocity results of a mobile robot in a fusion algorithm test provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0052] In this embodiment, a mobile robot path planning method based on an improved adaptive dynamic window method is provided. Figure 1 As shown, the following steps are included:

[0053] Step 1: Collect the robot's working environment information and establish a working environment map of the mobile robot;

[0054] In this embodiment, the working environment information of the mobile robot is collected through the visual camera, laser radar sensor, ultrasonic sensor, and infrared sensor equipped by the mobile robot, and then the working environment map of the mobile robot is established by using information fusion and SLAM technology;

[0055] Step 2: Set the starting point and target point of the mobile robot;

[0056] Set the current position of the mobile robot as the starting point Q start , the task location of the mobile robot is the target point Q goal ;

[0057] Step 3: Divide the working environment map Map of the mobile robot into m×n grids, and determine the starting point grid and the target point grid;

[0058] In this embodiment, the working environment map Map of the mobile robot is divided into m×n grids represented by Map={R(i,j)}, where R(i,j) represents the grid located in the i-th row and j-th column of the map Map, where i=0,1,…,m,j=0,1,…,n, and then the starting point Q start The grid where it is located is marked as the starting grid R(X start ,Y start ), the target point Q goal The grid where it is located is marked as the target grid R(X goal ,Y goal ), where X start , Y start They represent the horizontal and vertical coordinates of the starting grid on the map, X goal , Y goal Respectively represent the horizontal and vertical coordinates of the target grid on the map;

[0059] Step 4: Take the starting point as the parent node, expand the domain child nodes of the parent node and put them into the list of alternative path nodes (Openlist);

[0060] Step 5: Use the improved cost function in the A-star algorithm to calculate the cost of the domain sub-nodes in the candidate path node list (Openlist), select the sub-node with the minimum cost according to the cost calculation and move it into the path node list (Closelist) of the A-star algorithm;

[0061] The cost function of the A-star algorithm is improved and the cost is calculated using the improved cost function in the A-star algorithm. The improved cost function is shown in formula (1):

[0062] f(n)=g(n)+w(n)·h(n) (1)

[0063] Where n is the current node position; f(n) is the cost function, which represents the estimated cost from the current node n to the destination (total cost of the journey); g(n) is the actual cost from the starting point to the current node n; h(n) is the heuristic function, which represents the estimated cost from the current node n to the destination; w(n) is the weight function of the heuristic function h(n);

[0064] The calculation of the weight function is shown in formula (2):

[0065]

[0066] Among them, D star (n) is the distance from the starting point to the current node n, D goal (n) is the distance from the current node n to the target point, α and β are parameters that can be adjusted according to the actual planning area;

[0067] Step 6: Determine whether the current node has reached the target point Q goal If the target point Q is not reached goal , then take the current node as the parent node and expand the domain child nodes of the parent node, put them into the list of alternative path nodes (Openlist), and repeat steps 5 to 6 until the target point is reached;

[0068] Step 7: Reversely traverse the nodes in the path node list (Closelist) of the A-star algorithm to obtain a path from the starting point Q start To the target point Q goal The global optimal path of

[0069] Step 8: Optimize the global optimal path by removing redundant nodes to obtain an optimized path without redundant nodes;

[0070] Step 8.1: Group the nodes in the global optimal path and determine whether each group of nodes is on the same straight line. If so, skip the middle node in each group of nodes to simplify the path.

[0071] In this embodiment, each node is used as the step length of the loop. Three consecutive nodes are selected each time and connected in sequence to form two line segments as a group for looping. The vector cross product value of the two line segments is calculated to determine whether the three points are on the same straight line. If the three nodes are on the same straight line, the subnodes between the three nodes are skipped, thereby simplifying the path.

[0072] Step 8.2: If each group of nodes is not on the same straight line, calculate the equation of the straight line between the nodes to determine whether there is an obstacle on the straight line. If there is no obstacle, directly connect the first node and the last node to simplify the path. Repeat step 8.1 to continuously detect nodes that meet the conditions for optimization;

[0073] If the three nodes in a detected group are not on the same straight line, a redundancy removal algorithm is used to calculate the straight line equation between the three nodes, and the obstacle coordinate list is obtained to determine whether there is an obstacle on the straight line. If there is no obstacle, the first node and the third node are directly connected to simplify the redundant turning path;

[0074] Step 9: Perform arc interpolation processing on the turning corners in the optimized path obtained in step 8 to generate an optimized path;

[0075] Step 9.1: Loop through the coordinates of the nodes in the optimized path, perform arc interpolation processing on the turning path in the optimized path, and obtain an arc path;

[0076] The remaining paths that cannot be optimized after the optimization in step 8 are all inevitable turning paths. In the programming program, because the overall path has changed after the optimization in step 8, the variables storing the node coordinates need to be updated; loop with each node as the loop step length, and perform arc interpolation processing after obtaining the node coordinates; set the three adjacent nodes that are not on the same straight line as A1, A2, and A3, respectively. These three nodes form a corner, and connect the three adjacent nodes in turn to form two line segments A1A2 and A2A3, and compare the lengths of the two line segments. Take the endpoint of the short line segment as p and the endpoint of the long line segment as q; if A1A2 is a short line segment, then A1 is the endpoint p of the short line segment, and A3 is the endpoint q of the long line segment; if A2A3 is a shorter line segment, then the node A3 is the endpoint p of the short line segment, and A1 is the endpoint q of the long line segment; the center of the constructed arc will be located at the intersection of the perpendicular bisectors of the two line segments pA2 and A2q. The coordinates of the center of the circle can be solved by the simultaneous equations of the intersection of the two perpendicular lines. The radius of the arc can be obtained through the coordinates of the center of the circle. The arc point set generated by the center of the circle and the arc radius is connected to obtain the smoothed arc path.

[0077] Step 9.2: Check whether the obstacle exists within the arc range. If the generated arc path collides with the obstacle, adjust the end point position of the line segment to generate a new arc path;

[0078] By traversing the obstacle position matrix, it is determined whether the arc path conflicts with the obstacle and whether the obstacle exists within the arc range; if the generated arc collides with the obstacle, a new arc is generated by adjusting the position of the segment line endpoint p, that is, shrinking the arc inward until an obstacle-free path is generated.

[0079] Step 10: Segment the new arc path after the arc interpolation processing in step 9 to generate local target points;

[0080] Step 11: Use the local target points generated in step 10 as the multi-segment local targets of the improved adaptive dynamic window method, generate a speed combination based on the target point position, perform speed sampling, and generate multiple simulated motion trajectories for the next time node;

[0081] The speed sampling stage is to divide the speed window into multiple discrete speed values ​​to generate multiple possible linear speed and angular speed combinations;

[0082] Step 12: Select the optimal path according to the evaluation function and deflection screening mechanism of the improved adaptive dynamic window method;

[0083] Step 12.1: By improving the weight of the speed evaluation function, the evaluation function in the adaptive dynamic window method is improved, and the path of the mobile robot is dynamically adjusted using the improved evaluation function. The improved evaluation function is shown in formula (3):

[0084] G(v,ω)=σ(α*heading(v,ω)+β*dist(v,ω)+γd*vel(v,ω)) (3)

[0085] Among them, G(v,ω) is the improved evaluation function, σ is the normalization processing, α is the weight of the heading angle evaluation function, heading(v,ω) is the heading angle evaluation function, β is the weight of the obstacle distance evaluation function, dist(v,ω) is the obstacle distance evaluation function, γd is the weight of the improved speed evaluation function, vel(v,ω) is the speed evaluation function, v is the linear velocity of the mobile robot, and ω is the angular velocity of the mobile robot;

[0086] The weight calculation of the improved speed evaluation function is shown in formulas (4) and (5):

[0087]

[0088]

[0089] Among them, γd is set to a weight range of (0.1 to 0.4) in this implementation; Dist is the minimum obstacle distance; ε and is an adjustable parameter. In this embodiment, ε is set to 0.4. The value of x is 0.1; d is the horizontal coordinate of the nearest obstacle, x n is the horizontal coordinate of the mobile robot, y d is the ordinate of the nearest obstacle, y n is the vertical coordinate of the mobile robot; R is the obstacle radius, which is set to 0.5 in this embodiment;

[0090] By using the weight of the improved speed evaluation function, the weight of the speed evaluation function in the evaluation function formula (3) is modified so that the weight of the speed evaluation function is bound to the distance of the obstacle. The dynamic path adjustment capability of the mobile robot is realized through the above improvement.

[0091] Step 12.2: Use the dynamic weight judgment mechanism of deflection degree screening to assist in adjusting the weight of the speed evaluation function;

[0092] In the programming of this embodiment, the deflection degree screening mechanism is equivalent to a trigger mechanism, which stops the error weight of the previous step from continuing.

[0093] In the dynamic adjustment of the path, a dynamic weight judgment mechanism for deflection screening is added to calculate the deviation of the orientation angle between the nearest obstacle and the mobile robot; the orientation angle of the mobile robot is calculated by calling the toDegree function and converted into an angle value, and the azimuth angle of the mobile robot to the nearest obstacle is calculated and converted into an angle value, and the two angle values ​​obtained above are compared to obtain the direction difference between the robot and the nearest obstacle; if the direction difference exceeds 90°, it indicates that the mobile robot and the nearest obstacle are facing away from each other or the deflection is large (the obstacle is behind the robot), and the nearest obstacle will not affect the evaluation result of the subsequent mobile robot trajectory and will not cause movement obstacles to the robot. The mobile robot should resume the high-speed motion state and no longer continue to dynamically adjust the path of the mobile robot according to the nearest obstacle in step 12.1, ignoring the weight generated by the obstacle until the nearest obstacle to the robot is changed. This mechanism re-enters the waiting state for triggering, that is, when the mobile robot approaches the next nearest obstacle with a deflection difference of less than 90°, step 12.1 is performed again, and the path of the mobile robot is dynamically adjusted. The addition of the judgment mechanism reduces the influence of obstacles in complex terrain on the overall evaluation function.

[0094] Step 12.3: Select the optimal path according to the evaluation function and deflection screening mechanism of the improved adaptive dynamic window method;

[0095] Step 12.4: The mobile robot moves according to the optimal path to determine whether the mobile robot reaches the target point Q goal If the target point Q is not reached goal Repeat steps 11 to 12 until the robot reaches the target point Q according to the optimal path. goal ;

[0096] Step 13: After step 12, a complete optimized path planning and simulated motion route are obtained.

[0097] Figure 2 The upper path is obtained by using the improved A-star algorithm, and the lower path is obtained by using the traditional A-star algorithm. It can be seen from the figure that the path after the improved A-star algorithm is smoother and fewer nodes are traversed than the traditional A-star algorithm. It can be seen from Table 1 that the improved A-star algorithm adopted in the present invention has a faster search speed than the traditional A-star algorithm.

[0098] Table 1 Comparison between traditional A-star algorithm and improved A-star algorithm

[0099]

[0100] In order to verify the effectiveness and feasibility of the improved adaptive dynamic window method adopted in the present invention, a simulation is performed using matlab in an Intel Corei7-12700H processor with a main frequency of 2.3Ghz and a memory of 32GB. Figure 3 As shown, the map is a two-dimensional plane of 17m×17m. The circle in the figure is a circular obstacle with a diameter of 1m. The starting point of the path is the coordinate (0,0), and the target point is the coordinate (10,10).

[0101] Figure 4 Figures (a) and (b) are the simulation results of the traditional adaptive dynamic window method (DWA), and Figures (c) and (d) are the simulation results of the improved adaptive dynamic window method adopted by the present invention. Table 2 compares the running time of the traditional adaptive dynamic window method (DWA) and the improved adaptive dynamic window method adopted by the present invention.

[0102] Table 2 Comparison of running time between traditional adaptive dynamic window method (DWA) and improved adaptive dynamic window method

[0103]

[0104] from Figure 4As can be seen from Table 2, the improved adaptive dynamic window method adopted in the present invention is smoother than the traditional adaptive dynamic window method (DWA) in the speed scheduling of the mobile robot in the face of narrow road conditions, thereby improving the overall operation speed of the mobile robot and effectively reducing the path planning time.

[0105] Figure 5 The test result diagram of the improved A-star algorithm and the improved adaptive dynamic window method fusion algorithm in the present invention is shown in the figure. The environment map is a two-dimensional plane space of 30m×30m. The black squares in the figure are obstacles, the gray boxes are unknown temporary obstacles, and the rest of the space is movable space. The upper left triangle icon is the starting point of the mobile robot, and the circle icon below is the path planning target point. The thick solid line path is the running trajectory of the mobile robot, and the dotted line is the global path planned by the improved A-star algorithm. The mobile obstacle moves in a uniform straight line, and the straight line trajectory is the obstacle moving path. During the test, the mobile robot actively slows down to avoid collision when it intersects with the mobile obstacle, and increases the speed to the target point at the end of the intersection. There is no collision on the entire path, and it can avoid dynamic moving obstacles and has stability. Figure 6 This is a graph of the linear velocity and angular velocity of the mobile robot in the fusion algorithm test.

[0106] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope defined by the claims of the present invention.

Claims

1. A mobile robot path planning method based on an improved adaptive dynamic window method, characterized in that: The following steps are involved: Step 1: Collect the robot's working environment information and establish a working environment map of the mobile robot; Step 2: Set the starting point and target point of the mobile robot; Step 3: Divide the working environment map of the mobile robot into grids, and determine the starting point grid and the target point grid; Step 4: Take the starting point as the parent node, expand the domain child nodes of the parent node and put them into the list of candidate path nodes; Step 5: Use the improved cost function in the A-star algorithm to calculate the cost of the domain sub-nodes in the candidate path node list, select the sub-node with the minimum cost according to the cost calculation and move it into the path node list of the A-star algorithm; Step 6: Determine whether the current node has reached the target point. If not, take the current node as the parent node and expand the domain child nodes of the parent node, put them into the list of candidate path nodes, and repeat steps 5 to 6 until the target point is reached; Step 7: Reversely traverse the nodes in the path node list of the A-star algorithm to obtain a global optimal path from the starting point to the target point; Step 8: Optimize the global optimal path by removing redundant nodes to obtain an optimized path without redundant nodes; Step 9: Perform arc interpolation processing on the turning corners in the optimized path obtained in step 8 to generate an optimized path; Step 10: Segment the new arc path after the arc interpolation processing in step 9 to generate local target points; Step 11: Use the local target points generated in step 10 as the multi-segment local targets of the improved adaptive dynamic window method, generate a speed combination based on the target point position, perform speed sampling, and generate multiple simulated motion trajectories for the next time node; Step 12: Select the optimal path according to the evaluation function and deflection screening mechanism of the improved adaptive dynamic window method; Step 13: Get the complete optimized path planning and simulated motion line.

2. The mobile robot path planning method based on the improved adaptive dynamic window method according to claim 1 is characterized in that: The specific method of step 5 is: The cost function of the A-star algorithm is improved, and the improved cost function in the A-star algorithm is used for cost calculation. The improved cost function is shown in the following formula: f(n)=g(n)+w(n)·h(n) Where n is the current node position; f(n) is the cost function; g(n) is the actual cost of reaching the current node n from the starting point; h(n) is the heuristic function; w(n) is the weight function of the heuristic function h(n); The calculation of the weight function is shown in the following formula: Among them, D star (n) is the distance from the starting point to the current node n, D goal (n) is the distance from the current node n to the target point, and α and β are parameters that can be adjusted according to the actual planning area.

3. The mobile robot path planning method based on the improved adaptive dynamic window method according to claim 2 is characterized in that: The specific method of step 8 is: Step 8.1: Group the nodes in the global optimal path and determine whether each group of nodes is on the same straight line. If so, skip the middle node in each group of nodes to simplify the path. Step 8.2: If each group of nodes is not on the same straight line, calculate the equation of the straight line between the nodes and determine whether there are obstacles on the straight line. If there are no obstacles, directly connect the first node and the last node to simplify the path. Repeat step 8.1 to continuously detect nodes that meet the conditions for optimization.

4. The mobile robot path planning method based on the improved adaptive dynamic window method according to claim 3 is characterized in that: The specific method of step 9 is: Step 9.1: Loop through the coordinates of the nodes in the optimized path, perform arc interpolation processing on the turning path in the optimized path, and obtain an arc path; Step 9.2: Check whether the obstacle exists within the arc range. If the generated arc path collides with the obstacle, adjust the position of the line segment endpoints to generate a new arc path.

5. The mobile robot path planning method based on the improved adaptive dynamic window method according to claim 4 is characterized in that: The specific method of step 12 is: Step 12.1: Improve the evaluation function in the adaptive dynamic window method by improving the weight of the speed evaluation function, and dynamically adjust the path of the mobile robot using the improved evaluation function; Step 12.2: Use the dynamic weight judgment mechanism of deflection degree screening to assist in adjusting the weight of the speed evaluation function; Step 12.3: Select the optimal path according to the evaluation function and deflection screening mechanism of the improved adaptive dynamic window method; Step 12.4: The mobile robot moves according to the optimal path, and determines whether the mobile robot reaches the target point. If it does not reach the target point, repeat steps 11 to 12 until the robot moves according to the optimal path and reaches the target point.

6. The mobile robot path planning method based on the improved adaptive dynamic window method according to claim 5 is characterized in that: The improved evaluation function described in step 12.1 is shown in the following formula: G(v,ω)=σ(α*heading(v,ω)+β*dist(v,ω)+γd*vel(v,ω)) Among them, G(v,ω) is the improved evaluation function, σ is the normalization processing, α is the weight of the heading angle evaluation function, heading(v,ω) is the heading angle evaluation function, β is the weight of the obstacle distance evaluation function, dist(v,ω) is the obstacle distance evaluation function, γd is the weight of the improved speed evaluation function, vel(v,ω) is the speed evaluation function, v is the linear velocity of the mobile robot, and ω is the angular velocity of the mobile robot; The weight calculation of the improved speed evaluation function is shown in the following formula: Where Dist is the minimum obstacle distance; ε and is an adjustable parameter; x d is the horizontal coordinate of the nearest obstacle, x n is the horizontal coordinate of the mobile robot, y d is the ordinate of the nearest obstacle, y n is the vertical coordinate of the mobile robot; R is the radius of the obstacle.

7. The mobile robot path planning method based on the improved adaptive dynamic window method according to claim 6 is characterized in that: The specific method of step 12.2 is: adding a dynamic weight judgment mechanism for deflection screening in the dynamic adjustment of the path, calculating the deviation of the orientation angle between the nearest obstacle and the mobile robot; calculating the orientation angle of the mobile robot and converting it into an angle value, calculating the azimuth angle of the mobile robot to the nearest obstacle and converting it into an angle value, comparing the two angle values ​​obtained above to obtain the direction difference between the robot and the nearest obstacle; if the direction difference exceeds 90°, the mobile robot resumes high-speed motion and no longer continues to dynamically adjust the path of the mobile robot according to the nearest obstacle in step 12.

1. When the mobile robot approaches the next nearest obstacle with a deflection difference less than 90°, step 12.1 is performed again, and the path of the mobile robot is dynamically adjusted again.

Images