A path planning method and system for adaptive tangential obstacle avoidance

Abstract

The present invention discloses a path planning method and system for adaptive tangential obstacle avoidance. The system includes a laser radar, a Raspberry Pi, an STM32 controller, a motor, and a motor drive. The method utilizes a fuzzy controller to adaptively adjust a repulsion gain coefficient, thereby improving a robot's adaptability to changes in the surrounding environment and accelerating the robot's approach to a target point when near the target point. The method evaluates the importance of obstacles influencing the robot's movement path and classifies them into important obstacles and ordinary obstacles. Only important obstacles need to be considered during tangential obstacle avoidance, thereby reducing computational complexity. The repulsive influence range of important obstacles is divided into a safe distance area, a tangential obstacle avoidance area, and an artificial potential field area, thereby improving the safety of the robot's obstacle avoidance. The fuzzy controller is used for tangential obstacle avoidance, thereby avoiding the problem of local minimum values. At the same time, the method avoids drastic changes in the direction of the resultant force in narrow channels between obstacles, thereby avoiding path oscillation.

Description

Technical Field

[0001] The present invention relates to a path planning method and system for adaptive tangential obstacle avoidance, belonging to the technical field of path planning. Background Art

[0002] Mobile robots move toward their destinations based on a global path planning algorithm, such as the A* algorithm. However, the A* algorithm can only plan paths based on a global map. Furthermore, when the external environment changes, the A* algorithm cannot perceive these changes, causing the robot to collide with obstacles and path planning failure. Therefore, a local path planning algorithm that can perceive environmental changes is needed to complement the global path planning algorithm to achieve obstacle avoidance.

[0003] Among the currently popular local path planning algorithms for mobile robots, the dynamic window method has a complex model and lacks foresight, resulting in long planned paths and limited real-time obstacle avoidance performance. The time elastic band method requires the distance, angle, and time differences between two states to calculate the velocity and angular velocity within the control cycle, resulting in high computational complexity and unstable control. The artificial potential field method, however, has been widely used due to its simple principle, good real-time performance, and low computational complexity. However, traditional artificial potential field methods face challenges such as target unreachability, local minima, and path oscillation, and cannot effectively perform path planning tasks in complex environments. For robots moving in complex environments, existing local path planning algorithms include the dynamic window method, the time elastic band method, and the artificial potential field method. The dynamic window method lacks foresight, and the time elastic band method suffers from unstable control. Both algorithms also have high model and computational complexity. While the artificial potential field method has a simple principle, good real-time performance, and low computational complexity, it also suffers from drawbacks such as target unreachability, local minima, and path oscillation. Summary of the Invention

[0004] In response to the problems existing in the above-mentioned prior art, the present invention provides a path planning method and system for adaptive tangential obstacle avoidance, which solves the problem of unreachable targets by improving the repulsive potential field function of the traditional artificial potential field method; through an adaptive repulsive gain coefficient fuzzy controller, the repulsive gain coefficient changes with changes in the surrounding environment, thereby improving the robot's adaptability to changes in the surrounding environment; through the tangential obstacle avoidance fuzzy controller, the local minimum problem and the path oscillation phenomenon in narrow channels between obstacles are avoided.

[0005] In order to achieve the above object, the technical solution adopted by the present invention is: a path planning system for adaptive tangential obstacle avoidance, the system includes a laser radar, a Raspberry Pi, an STM32 controller, a motor and a motor driver;

[0006] The laser radar is used to collect environmental information and communicate with the Raspberry Pi through a USB interface to transmit the laser radar data to the map construction unit;

[0007] The STM32 controller is used to control the actuator, transmit information with the Raspberry Pi via the serial port, and complete task scheduling such as acquisition speed and motion control;

[0008] The Raspberry Pi is used for global path planning and local path planning; sensor data is mapped and modeled through a map building unit to generate a global cost map and a local cost map; after setting the target point, the global path planning unit completes the global path planning and transmits it to the global path correction unit; the local path planning first calculates the distance between the robot and the obstacle, the angle between the resultant force and the obstacle, and the angle between gravity and the tangent through a distance calculation unit and an angle calculation unit and fuzzifies them, then performs fuzzy reasoning and defuzzification based on a fuzzy rule library, and finally completes the local path planning through a repulsion gain coefficient adaptive adjustment unit and an adaptive tangential obstacle avoidance unit and transmits it to the correction unit for global path correction.

[0009] A path planning method for adaptive tangential obstacle avoidance includes the following steps:

[0010] Step 1: Add the distance factor between the robot and the target point to the traditional repulsion function to enhance the attraction and weaken the repulsion; the improved repulsion consists of two parts, the force F rep1 The force F is directed from the obstacle to the robot. rep2 The robot points to the target point;

[0011] Step 2: The adaptive repulsion gain coefficient fuzzy controller takes the distance between the robot and the obstacle and the distance between the robot and the target point as input and adaptively adjusts the repulsion gain coefficient according to the fuzzy rule;

[0012] Step 3: Assess the importance of obstacles by determining whether they affect the robot's movement direction. These obstacles are classified as important or common. An important obstacle is defined as one in which the angle ω between the robot's combined force and the obstacle is small, and the distance d between the robot and the obstacle is close or moderate. A common obstacle is defined as one in which the angle ω between the robot's combined force and the obstacle is large, or the distance d between the robot and the obstacle is far.

[0013] Step 4: For important obstacles, divide the repulsive force influence range into a safe distance area, a tangential obstacle avoidance area, and an artificial potential field area;

[0014] Step 5: For important obstacles, the robot will perform tangential obstacle avoidance after entering the tangential obstacle avoidance area;

[0015] Step 6: The adaptive tangential obstacle avoidance fuzzy controller is designed with three inputs and one output, which takes the distance d between the robot and the obstacle, the angle ω between the resultant force and the obstacle, and the gravitational force and the adjusted repulsive force F as inputs. rep1The angle δ is used as input, and the robot will judge whether to be controlled by the artificial potential field, avoid obstacles in the tangential direction, or perform emergency braking in the potential field according to the output value output;

[0016] Step 7: Finally, the robot's movement mode is determined based on the range of the fuzzy controller output value.

[0017] The specific steps of step 5 are as follows: first, connect the center of the robot and the center of the obstacle with a straight line, and generate an intersection point with the boundary of the obstacle; calculate a straight line passing through the intersection point and tangent to the obstacle, and calculate the repulsive force component F at this time. rep1 Adjust it to be parallel to the direction of the straight line. The direction is determined by the positive or negative angle ω. Define the clockwise direction as a positive angle and the counterclockwise direction as a negative angle. When the angle between the resultant force and the obstacle is positive, the important obstacle is located on the right side of the robot. rep1 points to the left along the parallel line; when the angle between the resultant force and the obstacle is negative, the important obstacle is on the left side of the robot, F rep1 Pointing to the right along a parallel line;

[0018] For the adjusted repulsive force F rep1 , if it is related to the gravitational force F att The angle δ is small, and the robot maintains the control of the artificial potential field; if the angle δ is small, ... att When the angle δ is large, the robot starts to avoid obstacles in the tangential direction and moves one step in the direction of the current resultant force. After moving one step, the robot adjusts the direction of the repulsive force component in the same way and continues to move one step until the angle δ is small. Then, the robot exits the tangential obstacle avoidance and resumes the artificial potential field control.

[0019] The beneficial effects of the present invention are as follows: the present invention has an improved design. By improving the repulsive potential field function of the traditional artificial potential field method, the problem of target unreachability is solved; an adaptive repulsive gain coefficient fuzzy controller is used to adjust the repulsive gain coefficient in accordance with changes in the surrounding environment, thereby improving the robot's adaptability to changes in the surrounding environment; and a tangential obstacle avoidance fuzzy controller is used to avoid local minima and path oscillation in narrow passages between obstacles. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a force analysis diagram of the improved repulsive force function of the present invention;

[0021] Figure 2 This is a schematic diagram of obstacle importance assessment according to the present invention;

[0022] Figure 3 This is a schematic diagram of the division of the area affected by the repulsive force of important obstacles of the present invention;

[0023] Figure 4 A schematic diagram of the robot of the present invention performing tangential obstacle avoidance;

[0024] Figure 5 This is a schematic diagram of the robot exiting tangential obstacle avoidance according to the present invention;

[0025] Figure 6 This is a schematic diagram of the robot's tangential obstacle avoidance process of the present invention;

[0026] Figure 7 Schematic diagram of the path planning system of the present invention;

[0027] Figure 8 Schematic diagrams of different experimental environments for verifying the effectiveness of the present invention. Figure (a) shows the artificial potential field method after improving the repulsion function, Figure (b) shows the artificial potential field method after adaptively adjusting the repulsion gain coefficient through the fuzzy controller, and Figure (c) shows the artificial potential field method after adaptive tangential obstacle avoidance.

[0028] Figure 9 The figures are comparison diagrams verifying the performance of the present invention under different environmental complexities. Figure (a) shows the path effect of the escape force method in a simple environment; Figure (b) shows the path effect of the tangential obstacle avoidance of the present invention in a simple environment; Figure (c) shows the path effect of the escape force method in a more complex environment; Figure (d) shows the path effect of the tangential obstacle avoidance of the present invention in a more complex environment; Figure (e) shows the path effect of the escape force method in a complex environment; and Figure (f) shows the path effect of the tangential obstacle avoidance of the present invention in a complex environment. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. However, it should be understood that the specific embodiments described herein are only used to illustrate the present invention and are not intended to limit the scope of the present invention.

[0030] Unless otherwise defined, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention pertains. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0031] The main purpose of the invention is to provide a path planning method for adaptive tangential obstacle avoidance for mobile robots in complex environments and a hybrid path planning system. This method aims to address the shortcomings of the artificial potential field method in local path planning while improving performance in complex environments.

[0032] To achieve the above objectives, the present invention provides an adaptive tangential obstacle avoidance method for a mobile robot. Tangential obstacle avoidance in the present invention involves continuously adjusting the direction of the repulsive force component of an obstacle to the tangent direction of the intersection of the line connecting the robot's center and the obstacle's center and the obstacle, thereby achieving obstacle avoidance by allowing the robot to perform a circular motion around the obstacle. The specific steps are as follows:

[0033] First, the distance factor between the robot and the target point is added to the traditional repulsion function to enhance the attraction and weaken the repulsion. The improved repulsion consists of two parts: the force F rep1 The force F is directed from the obstacle to the robot. rep2 The robot points to the target point, such as Figure 1 shown.

[0034] Secondly, the adaptive repulsion gain coefficient fuzzy controller takes the distance between the robot and the obstacle and the distance between the robot and the target point as input, and adaptively adjusts the repulsion gain coefficient according to fuzzy rules. The fuzzy rules are shown in Table 1 below.

[0035] Table 1 Fuzzy rules table of repulsion gain coefficient

[0036]

[0037] {HJ, JJ, SZ, JY, HY} in the first row of Table 1 is the fuzzy set of the distance between the robot and the obstacle, and the corresponding fuzzy linguistic variables are {very close, relatively close, moderate, relatively far, very far}; {HJ, JJ, SZ, JY, HY} in the first column is the fuzzy set of the distance between the robot and the target point, and the corresponding fuzzy linguistic variables are consistent with d1; NS, MD, and PB are three different repulsion gain coefficients, indicating that the repulsion gain coefficient is small, moderate, and large, respectively.

[0038] Secondly, the importance of obstacles is evaluated by whether they affect the robot's movement direction. Obstacles are divided into important obstacles and ordinary obstacles. Important obstacles are defined as: the angle ω between the robot's resultant force direction and the obstacle is small and the distance d between the robot and the obstacle is close or moderate; ordinary obstacles are defined as: the angle ω between the robot's resultant force direction and the obstacle is large or the distance d between the robot and the obstacle is far, such as Figure 2 shown.

[0039] For important obstacles, the repulsive force influence range is divided into safe distance area, tangential obstacle avoidance area, artificial potential field area, such as Figure 3 As shown in the figure, multiple circles constitute the range of influence of the obstacle repulsion. The circle outside the obstacle is the safe distance area. When the robot enters this area, it will take emergency braking to avoid collision with the obstacle. The circle outside the safe distance area is the tangential obstacle avoidance area. The robot continuously adjusts its position in this area. Figure 3 The repulsive force component F rep1 The outer ring of the tangential obstacle avoidance area is the artificial potential field area, and the robot maintains artificial potential field control in this area.

[0040] For important obstacles, the robot enters Figure 3After entering the tangential obstacle avoidance area, tangential obstacle avoidance will be performed. First, connect the robot center and the obstacle center with a straight line. This straight line will intersect with the obstacle boundary. Calculate the straight line that passes through the intersection and is tangent to the obstacle. The repulsive force component F at this time is rep1 Adjust it to be parallel to the direction of the straight line, and the direction is determined by the positive or negative of the included angle ω. This article defines clockwise as a positive angle and counterclockwise as a negative angle. When the angle between the resultant force and the obstacle is positive, the important obstacle is located on the right side of the robot, F rep1 points to the left along the parallel line; when the angle between the resultant force and the obstacle is negative, the important obstacle is on the left side of the robot, F rep1 Point to the right along the parallel line.

[0041] For the adjusted repulsive force F rep1 , if it is related to the gravitational force F att The angle δ is small, and the robot maintains the control of the artificial potential field; if the angle δ is small, ... att The angle δ is large, the robot starts to avoid obstacles in the tangential direction and moves one step in the direction of the current resultant force, such as Figure 4 shown.

[0042] After moving one step, the robot adjusts the direction of the repulsive force component in the same way and continues to move one step until the angle δ is small, then exits the tangential obstacle avoidance and resumes the artificial potential field control, as shown in the following example: Figure 5 shown.

[0043] The adaptive tangential obstacle avoidance fuzzy controller is designed with three inputs and one output, which takes the distance d between the robot and the obstacle, the angle ω between the resultant force and the obstacle, and the gravitational force and the adjusted repulsive force F as inputs. rep1 The angle δ is used as input. Based on the output value, the robot will determine whether to be controlled by the artificial potential field, avoid tangential obstacles, or perform emergency braking in the potential field. The semantic comparison of the fuzzy controller for tangential obstacle avoidance is shown in Table 2.

[0044] ■Table 2 Fuzzy semantic comparison table

[0045]

[0046] The following 10 fuzzy rules are designed for the tangential obstacle avoidance fuzzy controller:

[0047] a.If(d is DN)then(output is ZE);

[0048] b .If(d is DF)then(output is APF);

[0049] c.If(d is DM)and(ω is FB)then(output is APF);

[0050] d.If(d is DM)and(ω is ZB)then(output is APF);

[0051] e.If(d is DM)and(ω is FS)and(δis QB)then(output is RQM);

[0052] f.If(d is DM)and(ω is FS)and(δis QS)then(output is APF);

[0053] g.If(d is DM)and(ω is ZS)and(δis QB)then(output is LQM);

[0054] h.If(d is DM)and(ω is ZS)and(δis QS)then(output is APF);

[0055] i.If(d is DM)and(ω is ZJ)and(δis QB)then(output is LQM);

[0056] j.If(d is DM)and(ω is ZJ)and(δis QS)then(output is APF);

[0057] Finally, the robot's movement mode is determined according to the range of the fuzzy controller output value. The process is shown in Figure 6.

[0058] The present invention also provides a hybrid path planning system suitable for mobile robots, such as Figure 7 As shown in the figure, the system consists of a lidar, a Raspberry Pi, an STM32 controller, a motor, and a motor driver.

[0059] The lidar is used to collect environmental information and communicate with the Raspberry Pi through the USB port to transmit the lidar data to the map construction unit.

[0060] The STM32 controller is used to control the actuator, transmit information with the Raspberry Pi through the serial port, and complete task scheduling such as obtaining speed and motion control.

[0061] The Raspberry Pi is used for global and local path planning. Sensor data is mapped through the map building unit to generate global and local cost maps. After setting the target point, the global path planning unit completes global path planning and transmits it to the global path correction unit. Local path planning first uses the distance calculation unit and the angle calculation unit to calculate the distance between the robot and the obstacle, the angle between the resultant force and the obstacle, and the angle between gravity and the tangent, and then fuzzifies them. Fuzzy reasoning and defuzzification are then performed based on the fuzzy rule base. Finally, the repulsion gain coefficient adaptive adjustment unit and the adaptive tangential obstacle avoidance unit complete the local path planning and transmit it to the correction unit for global path correction.

[0062] Experimental verification: The rationality of the present invention is further illustrated by effectiveness verification. The experimental environment is a two-dimensional space of 60m*60m. The starting point and target point of the robot and the coordinates of the obstacle are set according to different experimental requirements. The gravitational coefficient of the traditional artificial potential field method is set to 60 by default, the repulsive coefficient is set to 50, the robot's moving step is set to 1m, the obstacle's influence distance is set to the obstacle radius plus 5m, and the NS, MD, and PB of the adaptive repulsive gain coefficient fuzzy controller are set to 10, 20, and 30 respectively. The magenta square is the starting point of the robot, the blue asterisk is the target point, the black circle is the obstacle, the dotted circle outside the black circle is the safe distance area, and the planned path is the red dot. Construct the following Figure 8 The experimental environment shown. Figure 8 (a) is the artificial potential field method after improving the repulsion function, Figure 8 (b) is the artificial potential field method after the fuzzy controller adaptively adjusts the repulsive force gain coefficient. Figure 8 (c) is the artificial potential field method after adaptive tangential obstacle avoidance. Figure 8 (a) Figure 8 The comparison with (b) shows that the artificial potential field method after adaptive adjustment makes the robot move forward a little, but because the collinearity situation is not changed, it still falls into the local minimum. Figure 8 (c) After adding tangential obstacle avoidance, the robot avoids the local minimum and reaches the target point smoothly, verifying the effectiveness of the present invention in solving the local minimum.

[0063] The present invention also verifies the performance under different environmental complexities. To this end, three experimental environments, simple, relatively complex, and complex, are constructed and compared with the artificial potential field method based on escape force. Figure 8 shown.

[0064] Depend on Figure 9 It can be seen that in simple and complex environments, the escape force method plans a longer path; in complex environments, the escape force method fails to successfully plan a path; when moving in a narrow channel between obstacles, the escape force method produces serious path oscillation, such as Figure 9(c) The total step length and time consumption for path planning for three environmental complexities are shown in Table 3. The results show that in both simple and complex environments, the proposed method achieves better path planning and takes less time than the escape force method. Even under complex conditions, the proposed method can still plan an optimal path and avoid obstacles.

[0065] ■Table 3 Comparative Analysis

[0066]

[0067] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions or improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A path planning method for adaptive tangential obstacle avoidance, characterized in that: The following steps are involved: Step 1: Add the distance factor between the robot and the target point to the traditional repulsion function to enhance the attraction and weaken the repulsion; the improved repulsion consists of two parts, the force F rep1 The force F is directed from the obstacle to the robot. rep2 The robot points to the target point; Step 2: The adaptive repulsion gain coefficient fuzzy controller takes the distance between the robot and the obstacle and the distance between the robot and the target point as input and adaptively adjusts the repulsion gain coefficient according to the fuzzy rule; Step 3: Evaluate the importance of obstacles by considering whether they affect the robot's movement direction, and classify them into important obstacles and ordinary obstacles. A significant obstacle is defined as one in which the angle ω between the robot's resultant force direction and the obstacle is small and the distance d between the robot and the obstacle is close or moderate. A common obstacle is defined as one in which the angle ω between the robot's resultant force direction and the obstacle is large or the distance d between the robot and the obstacle is far. Step 4: For important obstacles, divide the repulsive force influence range into a safe distance area, a tangential obstacle avoidance area, and an artificial potential field area; Step 5: For important obstacles, the robot will perform tangential obstacle avoidance after entering the tangential obstacle avoidance area; The specific steps of step 5 are as follows: first, connect the center of the robot and the center of the obstacle with a straight line L1, and generate an intersection between the straight line L1 and the boundary of the obstacle; calculate the straight line L2 passing through the intersection and tangent to the obstacle, and calculate the repulsive force F at this time. rep1 Adjust it to be parallel to the direction of the straight line L2. The direction is determined by the positive or negative angle ω between the robot's resultant force direction and the obstacle. Define the clockwise direction as a positive angle and the counterclockwise direction as a negative angle. When the angle between the robot's resultant force direction and the obstacle is positive, the important obstacle is located on the right side of the robot. rep1 Pointing to the left along the parallel line; when the angle between the robot's resultant force direction and the obstacle is negative, the important obstacle is located on the left side of the robot, F rep1 Pointing to the right along a parallel line; Step 6: The adaptive tangential obstacle avoidance fuzzy controller is designed with three inputs and one output, which takes the distance d between the robot and the obstacle, the angle ω between the robot's force direction and the obstacle, and the gravitational force F as input. att and the adjusted repulsive force F rep1 The angle δ is used as input, and the robot will judge whether to be controlled by the artificial potential field, avoid obstacles in the tangential direction, or perform emergency braking in the potential field according to the output value output; Step 7: Finally, the robot's movement mode is determined based on the range of the output value of the adaptive tangential obstacle avoidance fuzzy controller.

2. The path planning method for adaptive tangential obstacle avoidance according to claim 1, characterized in that: The step five also includes adjusting the repulsive force component F rep1 , if it is related to the gravitational force F att The angle δ is small, and the robot maintains the control of the artificial potential field; if the angle δ is small, ... att The angle δ is large, the robot starts to avoid obstacles in the tangential direction and moves one step in the direction of the current resultant force; after moving one step, the robot adjusts the repulsive force F in the same way rep1 direction and continue to move one step until the angle δ is small, then exit the tangential obstacle avoidance and resume artificial potential field control.

3. An adaptive tangential obstacle avoidance path planning system, which is applied to the adaptive tangential obstacle avoidance path planning method according to any one of claims 1 or 2, characterized in that: The system includes a lidar, a Raspberry Pi, an STM32 controller, a motor, and a motor driver; The laser radar is used to collect environmental information and communicate with the Raspberry Pi through a USB interface to transmit the laser radar data to the map construction unit; The STM32 controller is used to control the actuator, transmit information with the Raspberry Pi via the serial port, and complete the acquisition speed and motion control task scheduling; The Raspberry Pi is used for global path planning and local path planning; the sensor data is used for map modeling through the map building unit to generate a global cost map and a local cost map; after setting the target point, the global path planning unit completes the global path planning and transmits it to the global path correction unit; the local path planning first calculates the distance between the robot and the obstacle, the angle between the robot's combined force direction and the obstacle, and the gravitational force F through the distance calculation unit and the angle calculation unit. att and the adjusted repulsive force F rep1 The included angle is fuzzified, and then fuzzy reasoning and defuzzification are performed according to the fuzzy rule base. Finally, the local path planning is completed by the repulsion gain coefficient adaptive adjustment unit and the adaptive tangential obstacle avoidance unit and transmitted to the global path correction unit for global path correction.

Images