A LiDAR-based method for full-coverage path planning and obstacle avoidance in wall-climbing robots.

Abstract

This invention discloses a full-coverage path planning and obstacle avoidance system and method for wall-climbing robots based on lidar. The system, applied to a wall-climbing robot, includes a remote control module, a host computer, a lidar, and a chassis control module. The host computer, lidar, and chassis control module are all mounted on the wall-climbing robot. The remote control module provides a visual interface and control buttons for the operator. The remote control module sends control signals to the host computer and receives information from it. The host computer fuses and processes the received data and sends instructions to the chassis control module to control the wall-climbing robot's operation. This invention utilizes lidar for environmental perception of the wall-climbing robot, achieving full-coverage path walking and obstacle avoidance based solely on sidewall and ground features. The system has low complexity, improves the intelligence of the wall-climbing robot, and can replace manual remote control for flat tank wall operations, saving manpower and material costs.

Description

Technical Field

[0001] This invention relates to the field of robot control technology, specifically to a method for full-coverage path planning and obstacle avoidance for a wall-climbing robot based on lidar. Background Technology

[0002] As a fundamental industry in the nation, petrochemical enterprises rely heavily on storage tanks, which are essential equipment. To ensure the safe operation of petrochemical equipment, regular maintenance, including painting and rust removal, is necessary. Currently, traditional tank painting methods in China primarily involve manually erecting scaffolding, which suffers from long work cycles, low efficiency, health risks, and hazardous working environments.

[0003] With industrial upgrading, intelligent robot technology is being applied more and more in industrial production. For hazardous working environments on walls, using robots to replace manual labor is highly significant. Although magnetic wall-climbing robots are gradually being applied to such environments, they are still mainly operated remotely by humans, with low levels of automation and intelligence. Currently, there is a lack of mature and stable autonomous wall-climbing robots capable of full coverage.

[0004] In the field of environmental perception, LiDAR solutions offer advantages over vision-based solutions, including less susceptibility to external environmental influences, higher accuracy, and better stability. This makes them more suitable for harsh environments like those requiring all-around detection, such as wall-mounted structures. Processing data from a single LiDAR sensor is sufficient to meet the operational requirements of rectangular automated storage tanks, avoiding issues such as data redundancy, abnormal data crosstalk, and slow response times that arise from designing multiple sensor systems. Summary of the Invention

[0005] This invention proposes a full-coverage path planning and obstacle avoidance system and method for wall-climbing robots based on lidar, and proposes a new solution to the problems of high labor efficiency and low cost and low intelligence of wall-climbing robots in existing wall-climbing operation scenarios.

[0006] The technical solution of the present invention is as follows:

[0007] A LiDAR-based full-coverage path planning and obstacle avoidance system for wall-climbing robots is applied to wall-climbing robots, including a remote control module, a host computer, a LiDAR, and a chassis control module;

[0008] The lidar and remote control terminal belong to the data acquisition layer, the host computer belongs to the fusion decision layer, which is used to comprehensively judge environmental information and control terminal information to give motion commands, and the chassis control module belongs to the motion control layer, which is used to parse motion commands and control the movement of chassis motors.

[0009] The host computer, LiDAR, and chassis control module are all mounted on the wall-climbing robot. The LiDAR is used to scan the area around the robot in real time and transmit the point cloud information to the host computer for processing. The chassis control module is used to convert the commands issued by the host computer into drive signals for the motors through the kinematic model of the robot chassis, controlling the rotation of the motors to achieve the predetermined functions. The LiDAR is horizontally mounted on the back of the robot via a mounting bracket, with the X-axis of the LiDAR aligned with the front of the robot. The LiDAR collects environmental information around the wall-climbing robot in real time and generates point cloud data for processing.

[0010] The remote control module provides operators with a visual interface and control buttons for monitoring the status of the wall-climbing robot, remotely controlling the robot's movement and operations, setting initial map parameters, displaying the robot's navigation trajectory, setting the working arm's movement status, and controlling emergency stops.

[0011] The remote control module sends control signals to the host computer and receives information from the host computer. The host computer processes the received data and sends instructions to the chassis control module to control the wall-climbing robot's operation.

[0012] The aforementioned host computer is equipped with a lidar point cloud processing module, a full-coverage walking obstacle avoidance module, and an independent control module;

[0013] The lidar point cloud processing module is used to acquire the posture and relative position of the wall-climbing robot, acquire information about obstacles around the robot, and provide the data to the full-coverage walking obstacle avoidance module to realize its function.

[0014] The full-coverage walking obstacle avoidance module is used to comprehensively process information based on the instructions of the remote control module and the relevant information provided by the lidar point cloud processing module, and then control the chassis movement through the chassis control module to realize the robot's full-coverage walking obstacle avoidance function on the tank wall.

[0015] The independent control module is used to receive instructions from the remote control module to control the movement of the wall-climbing robot and the rotation of its working motors. Then, it informs the chassis control module through a communication protocol to realize the remote control operation of the wall-climbing robot.

[0016] The wall-climbing robot has two states: vertical movement and horizontal movement. The maximum distance of a single vertical movement is the set map height, and the maximum distance of a single horizontal movement is the movement distance set according to the robot's single working range.

[0017] A method for full-coverage path planning of a wall-climbing robot based on LiDAR, utilizing the aforementioned obstacle avoidance system, includes the following steps:

[0018] S1. After the wall-climbing robot is started, the lidar point cloud processing module works. It obtains point cloud information by scanning the robot's surrounding environment in real time, processes the point cloud information to obtain the robot's position information, attitude angle information and obstacle information, and sends the processing results to the full-coverage walking and obstacle avoidance module in real time.

[0019] S2, use the remote control module to set the width and height of the tank wall to be worked on, control the robot to move to the upper right corner of the tank wall to be worked on, and issue an automatic operation command. At this time, the wall-climbing robot automatically sets the target height.

[0020] S3. After receiving the automatic operation command, the wall-climbing robot enters the full-coverage walking and obstacle avoidance module, receives information sent from the lidar point cloud processing module, and makes real-time judgments on the information to make corresponding walking status decisions.

[0021] S4, the wall-climbing robot first moves vertically, that is, it moves automatically towards the target height. If a posture deviation is detected, proceed to step S5. If an obstacle is detected in the direction of movement, proceed to step S6. Then continue to move towards the target height. When it reaches the target height, switch the target height and proceed to step S8.

[0022] S5, When the wall-climbing robot's posture is greater than the posture threshold 1, the system performs posture correction logic, controls the robot's own rotation, and judges the robot's current posture in real time. When the posture is less than the posture threshold 2, the robot exits the posture correction logic.

[0023] S6, When moving vertically, if there is an obstacle in the direction of movement, the system enters the vertical obstacle avoidance logic. First, it switches to horizontal movement and simultaneously judges the obstacle information in both the vertical and horizontal directions. If there is no obstacle in the horizontal direction but an obstacle in the vertical direction, it moves horizontally until there is no obstacle in the vertical direction, and then continues to step S4. If the horizontal movement distance exceeds the maximum horizontal movement distance in a single step, it switches the target height and continues to step S4. If there are obstacles in both the horizontal and vertical directions, it switches the target height and continues to step S4.

[0024] S7. When moving horizontally, if there is an obstacle in the direction of movement, the system enters the horizontal obstacle avoidance logic. First, it switches to vertical movement and simultaneously judges the obstacle information in both the vertical and horizontal directions. If there is no obstacle in the vertical direction but an obstacle in the horizontal direction, it moves vertically until there is no obstacle in the horizontal direction, and then continues to step S8. If the vertical movement distance reaches the target height, it switches the target height and continues to step S8. If there are obstacles in both the horizontal and vertical directions, it switches the target height and continues to step S8.

[0025] S8, robot horizontal movement state. If posture deviation is detected, proceed to step S5; if there is an obstacle during horizontal movement, proceed to step S7; if the robot moves to the maximum distance of a single horizontal movement, proceed to step S4 and enter the vertical movement state.

[0026] S9. After the robot moves a total horizontal distance to the set width of the tank wall, the robot completes the full-coverage path planning and obstacle avoidance operation. The program exits the full-coverage path planning system, the progress independent control module is activated, the operation motor is turned off, and the robot is remotely controlled by the remote control module to perform the robot retrieval operation.

[0027] In step S1, the specific steps are as follows:

[0028] S1.1, Obtain the raw point cloud data from the LiDAR. Perform a pass-through filter on the raw data, retaining only the point cloud data of the parts of the ground and side walls that the LiDAR is facing when the wall-climbing robot is working. Then, use a planar segmentation fitting method based on Random Sample Consensus (RANSAC) to fit the side areas and the ground area respectively, fitting the ground and side wall surfaces, and storing the coefficients of the planar model. The linear equation of the planar model is as follows:

[0029] Ax + By + Cz + D = 0

[0030] Where A, B, and C represent the components of the plane normal vector, and D represents the distance from the plane to the origin. After obtaining the two plane models, the height information of the wall-climbing robot is calculated based on the distance from the origin of the lidar to the ground plane model.

[0031] The horizontal position information of the wall-climbing robot is calculated based on the distance from the origin of the lidar to the plane model of the side wall; the height information and the horizontal position information together constitute the position information of the robot on the tank wall.

[0032] S1.2, Since the LiDAR is mounted on the robot, the coordinate system of the LiDAR and the coordinate system of the robot are treated as the same coordinate system. After obtaining the planar models of the side wall and the ground, the vertical attitude angle of the robot is obtained based on the angle between the plane normal vectors of the two planes and the robot's x-axis. Considering the existence of plane fitting errors, the residuals δ1 and δ2 of the two fitted planes are calculated respectively. The attitude angle to be used by the wall-climbing robot program at the end is determined based on the magnitude of the error.

[0033]

[0034] Where θ1 and θ2 represent the robot's vertical attitude angles calculated from the side wall and the ground, respectively; δ1 and δ2 represent the residuals of the two planes fitted from the side wall and the ground, respectively; and θ represents the attitude angle of the wall-climbing robot used in the end. Obtaining the attitude angles gives the robot's attitude information.

[0035] S1.3, Copy a copy of the original LiDAR point cloud data, and perform obstacle detection processing on the copied original point cloud data. The processing method is to extend a certain distance outward according to the length and width of the robot to form a "+" shaped obstacle judgment range, and eliminate possible interference from itself. The extension distance is determined according to the obstacle avoidance threshold. When there are points of radar point cloud in the area, it can be said that there is an obstacle, and the direction of the obstacle is determined according to the coordinates of the points.

[0036] In step S4, the robot is constantly moving towards the target height during vertical movement. There are two target heights: one at the highest point of the map and one at the lowest point of the map. The highest point is the initial height at which the robot starts navigation in the upper right corner of the map, and the lowest point is the initial height minus the set map height. After the robot reaches a target height, the target height is switched, which is manifested by the robot moving up and down back and forth. Horizontal leftward movement is inserted between the upward and downward movements to achieve full-coverage path walking operation.

[0037] In step S5, there are two attitude thresholds. The attitude threshold 1 is larger and is used to determine whether the robot needs to correct its attitude to avoid excessive deviation of the work trajectory. The attitude threshold 2 is smaller and is used to correct the robot's attitude according to the smaller threshold when the robot is correcting its attitude, so as to ensure that the robot is corrected as accurately as possible.

[0038] The beneficial effects of this invention are:

[0039] 1. This invention utilizes lidar for environmental perception of the wall-climbing robot. It can achieve full-coverage path walking and obstacle avoidance with only side wall and ground features. The system has low complexity, improves the intelligence of the wall-climbing robot, and can replace manual remote control for flat tank wall operations, saving manpower and material costs.

[0040] 2. This invention uses a method based on two fitting planes to determine the attitude angle. The weight of the calculated attitude angle is determined according to the residual of the two planes, which makes the calculated robot attitude angle more accurate and improves the robot's operation accuracy and system stability.

[0041] 3. The present invention is a full-coverage path planning and obstacle avoidance system for wall-climbing robots based on lidar. By processing lidar point clouds to perceive the environment, it can complete the full-coverage path operation process of the wall-climbing robot on the square tank wall. At the same time, it has autonomous obstacle avoidance function, no need for SLAM mapping, low requirements for working environment, good system stability, and fills the gap in automated operation of square tank walls in the petrochemical industry. Attached Figure Description

[0042] Figure 1This is a system principle block diagram of a wall-climbing robot full-coverage path planning and obstacle avoidance system based on lidar according to the present invention;

[0043] Figure 2 This is a schematic diagram of the working environment of a wall-climbing robot full-coverage path planning and obstacle avoidance system based on lidar according to the present invention;

[0044] Figure 3 This is a modular design block diagram of a full-coverage path planning and obstacle avoidance system for a wall-climbing robot based on lidar, according to the present invention.

[0045] Figure 4 This is a schematic diagram illustrating the workflow of a LiDAR-based wall-climbing robot's full-coverage path planning and obstacle avoidance system according to the present invention. Detailed Implementation

[0046] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0047] A LiDAR-based full-coverage path planning and obstacle avoidance system for wall-climbing robots, such as... Figure 1 As shown, a host computer and chassis control module are installed inside the wall-climbing robot, while a LiDAR is installed externally. A remote control module is used for control. Signals from the LiDAR and remote control terminal are acquired, fused, and decided within the host computer. The decision result is then sent to the chassis control module for robot motion control.

[0048] The lidar and remote control terminal belong to the data acquisition layer, the host computer belongs to the fusion decision layer, which is used to comprehensively judge environmental information and control terminal information to give motion commands, and the chassis control module belongs to the motion control layer, which is used to parse motion commands and control the movement of chassis motors.

[0049] Working environment of wall-climbing robots, such as Figure 2 As shown, the lidar needs to scan the sides and the ground to fit a plane and determine its pose.

[0050] Modular design and control process of a LiDAR-based wall-climbing robot's full-coverage path planning and obstacle avoidance system, as follows: Figure 3 As shown, it includes a remote control module, a lidar point cloud processing module, a full-coverage walking obstacle avoidance module, an independent control module, and a chassis control module;

[0051] The remote control module is used for monitoring the status of the wall-climbing robot, remotely controlling the robot's movement and operation, setting initial map parameters, displaying the robot's navigation trajectory, setting the working arm's movement status, and controlling emergency stops.

[0052] The lidar point cloud processing module is used to acquire the posture and relative position of the wall-climbing robot, acquire information about obstacles around the robot, and provide the data to the full-coverage walking obstacle avoidance module to realize its function.

[0053] The full-coverage walking obstacle avoidance module is used to comprehensively process information based on the instructions of the remote control module and the relevant information provided by the lidar point cloud processing module, and then control the chassis movement through the chassis control module to realize the robot's full-coverage walking obstacle avoidance function on the tank wall.

[0054] The independent control module is used to receive instructions from the remote control module to control the movement of the wall-climbing robot and the rotation of its working motors. Then, it informs the chassis control module through a communication protocol to realize the remote control operation of the wall-climbing robot.

[0055] The chassis control module is used to convert the instructions issued by the host computer into drive signals for the motors through the kinematic model of the wall-climbing robot chassis, control the rotation of the motors, and achieve the predetermined functions.

[0056] The wall-climbing robot has two states: vertical movement and horizontal movement. The maximum distance of a single vertical movement is the set map height, and the maximum distance of a single horizontal movement is the movement distance set according to the robot's single working range.

[0057] A method for full-coverage path planning and obstacle avoidance for a wall-climbing robot based on LiDAR is illustrated in the following simplified flowchart: Figure 4 As shown, the steps are as follows:

[0058] S1. After the robot starts, the lidar point cloud processing module works to obtain point cloud information by scanning the robot's surrounding environment in real time.

[0059] S1.1, Perform a pass-through filter on the original data, retaining only the point cloud data of the parts of the wall-climbing robot facing the ground and side walls when the radar is working. Then, use a planar segmentation fitting method based on Random Sample Consensus (RANSAC) to fit the side regions and ground regions respectively, fitting the ground and side wall surfaces, and storing the coefficients of the planar model. The linear equation of the planar model is as follows:

[0060] Ax + By + Cz + D = 0

[0061] Where A, B, and C represent the components of the plane normal vector, and D represents the distance from the plane to the origin. After obtaining the two planar models, the height information of the wall-climbing robot is calculated based on the distance from the origin of the lidar to the ground planar model.

[0062] The horizontal position of the wall-climbing robot is calculated based on the distance from the origin of the lidar to the planar model of the side wall. The height information and the horizontal position information together constitute the robot's position information on the tank wall.

[0063] S1.2, Due to the mounting position of the LiDAR on the robot, the coordinate system of the LiDAR and the coordinate system of the robot can be approximated as the same coordinate system. After obtaining the planar models of the side walls and the ground, the vertical attitude angle of the robot is obtained based on the angle between the plane normal vectors of the two planes and the robot's x-axis. Considering the existence of plane fitting errors, the residuals δ1 and δ2 of the two fitted planes are calculated respectively. The attitude angle to be used by the wall-climbing robot program at the end is determined based on the magnitude of the error.

[0064]

[0065] Where θ1 and θ2 represent the robot's vertical attitude angles calculated from the side wall and the ground, respectively; δ1 and δ2 represent the residuals of the two planes fitted from the side wall and the ground, respectively; and θ represents the attitude angle of the wall-climbing robot used in the end. Obtaining the attitude angles gives the robot's attitude information.

[0066] S1.3, Copy a copy of the original LiDAR point cloud data, and perform obstacle detection processing on the copied original point cloud data. The processing method is to extend a certain distance outward according to the length and width of the robot to form a "+" shaped obstacle judgment range, and eliminate possible interference from itself. The extension distance is determined according to the obstacle avoidance threshold. When there are points of radar point cloud in the area, it can be said that there is an obstacle, and the direction of the obstacle is determined according to the coordinates of the points.

[0067] S2, use the remote control module to set the width and height of the tank wall to be worked on, control the robot to move to the upper right corner of the tank wall to be worked on, and issue an automatic operation command. At this time, the robot automatically sets the target height.

[0068] S3. After receiving the automatic operation command, the wall-climbing robot enters the full-coverage walking and obstacle avoidance module, receives information sent from the lidar point cloud processing module, makes real-time judgments on the information, and makes corresponding walking status decisions.

[0069] S4, the wall-climbing robot first moves vertically, that is, it moves automatically towards the target height. If a posture deviation is detected, proceed to step S5. If an obstacle is detected in the direction of movement, proceed to step S6. Then continue to move towards the target height. When it reaches the target height, switch the target height and proceed to step S8.

[0070] S5, When the wall-climbing robot's posture is greater than the posture threshold 1, the system performs posture correction logic, controls the robot's own rotation, and judges the robot's current posture in real time. When the posture is less than the posture threshold 2, the robot exits the posture correction logic.

[0071] S6, During vertical movement, when there is an obstacle in the direction of movement, the system enters the vertical obstacle avoidance logic. First, it switches to horizontal movement and simultaneously judges the obstacle information in both the vertical and horizontal movement directions. If there is no obstacle in the horizontal movement direction but an obstacle in the vertical movement direction, it moves horizontally until there is no obstacle in the vertical movement direction, then continues to step S4; if the horizontal movement distance exceeds the maximum horizontal movement distance in a single step, it switches the target height and continues to step S4; if there are obstacles in both the horizontal and vertical movement directions, it switches the target height and continues to step S4.

[0072] S7. When moving horizontally, if there is an obstacle in the direction of movement, the system enters the horizontal obstacle avoidance logic. First, it switches to vertical movement and simultaneously judges the obstacle information in both the vertical and horizontal directions. If there is no obstacle in the vertical direction but an obstacle in the horizontal direction, it moves vertically until there is no obstacle in the horizontal direction, then continues to step S8; if the vertical movement distance reaches the target height, it switches the target height and continues to step S8; if there are obstacles in both the horizontal and vertical directions, it switches the target height and continues to step S8.

[0073] S8, robot horizontal movement state. If posture deviation is detected, proceed to step S5; if there is an obstacle during horizontal movement, proceed to step S7; if the robot moves to the maximum distance of a single horizontal movement, proceed to step S4 and enter the vertical movement state.

[0074] S9. Once the robot's total horizontal movement distance reaches the set tank wall width, the robot completes the full-coverage path planning and obstacle avoidance operation. The program then exits the full-coverage path planning system, the progress independent control module is activated, and the operating motors are shut down. At this point, the robot can be remotely controlled using the remote control module for retrieval.

[0075] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for full-coverage path planning and obstacle avoidance of a wall-climbing robot based on lidar, utilizing a full-coverage path planning and obstacle avoidance system for a wall-climbing robot based on lidar, including a remote control module, a host computer, lidar, and a chassis control module; The lidar and remote control terminal belong to the data acquisition layer, the host computer belongs to the fusion decision layer, which is used to comprehensively judge environmental information and control terminal information to give motion commands, and the chassis control module belongs to the motion control layer, which is used to parse motion commands and control the movement of the chassis motor. The host computer, LiDAR, and chassis control module are all mounted on the wall-climbing robot. The LiDAR is used to scan the area around the robot in real time and transmit the point cloud information to the host computer for processing. The chassis control module is used to convert the commands issued by the host computer into drive signals for the motors through the kinematic model of the robot chassis, controlling the rotation of the motors to achieve the predetermined functions. The LiDAR is horizontally mounted on the back of the robot via a mounting bracket, with the X-axis of the LiDAR aligned with the front of the robot. The LiDAR collects environmental information around the wall-climbing robot in real time and generates point cloud data for processing. The remote control module provides operators with a visual interface and control buttons for monitoring the status of the wall-climbing robot, remotely controlling the robot's movement and operations, setting initial map parameters, displaying the robot's navigation trajectory, setting the working arm's movement status, and controlling emergency stops. The remote control module sends control signals to the host computer and receives information from the host computer. The host computer processes the received data and sends instructions to the chassis control module to control the wall-climbing robot to operate. Its features are, Includes the following steps: S1. After the wall-climbing robot is started, the lidar point cloud processing module works. It obtains point cloud information by scanning the robot's surrounding environment in real time, processes the point cloud information to obtain the robot's position information, attitude angle information and obstacle information, and sends the processing results to the full-coverage walking and obstacle avoidance module in real time. S1.1, Obtain the raw point cloud data from the LiDAR. Perform a pass-through filter on the raw data, retaining only the point cloud data of the parts of the ground and side walls that the wall-climbing robot is facing when it is working. Then, use a plane segmentation fitting method based on random sampling consistency to fit the side area and the ground area respectively, fitting the ground and side wall surfaces, and storing the coefficients of the plane model. The linear equation of the plane model is as follows: Ax + By + Cz + D = 0 Where A, B, and C represent the components of the plane normal vector, and D represents the distance from the plane to the origin. After obtaining the two plane models, the height information of the wall-climbing robot is calculated based on the distance from the origin of the lidar to the ground plane model. The horizontal position information of the wall-climbing robot is calculated based on the distance from the origin of the lidar to the plane model of the side wall; the height information and the horizontal position information together constitute the position information of the robot on the tank wall. S1.2, Since the LiDAR is mounted on the robot, the coordinate system of the LiDAR and the coordinate system of the robot are treated as the same coordinate system. After obtaining the planar models of the side wall and the ground, the vertical attitude angle of the robot is obtained based on the angle between the plane normal vectors of the two planes and the robot's x-axis. Considering the existence of plane fitting errors, the residuals δ1 and δ2 of the two fitted planes are calculated respectively. The attitude angle to be used by the wall-climbing robot program at the end is determined based on the magnitude of the error. Where θ1 and θ2 represent the robot's vertical attitude angles calculated from the side wall and the ground, respectively; δ1 and δ2 represent the residuals of the two planes fitted from the side wall and the ground, respectively; and θ represents the attitude angle of the wall-climbing robot used in the end. Obtaining the attitude angles gives the robot's attitude information. S1.3, Copy a copy of the original point cloud data of the LiDAR, and perform obstacle detection processing on the copied original point cloud data. The processing method is to extend a certain distance outward according to the length and width of the robot to form a "+" shaped obstacle judgment range, and eliminate possible interference from itself. The extension distance is determined according to the obstacle avoidance threshold. When there are points of the radar point cloud in the area, it can be said that there is an obstacle, and the direction of the obstacle is determined according to the coordinates of the point. S2, use the remote control module to set the width and height of the tank wall to be worked on, control the robot to move to the upper right corner of the tank wall to be worked on, and issue an automatic operation command. At this time, the wall-climbing robot automatically sets the target height. S3. After receiving the automatic operation command, the wall-climbing robot enters the full-coverage walking and obstacle avoidance module, receives information sent from the lidar point cloud processing module, and makes real-time judgments on the information to make corresponding walking status decisions. S4, the wall-climbing robot first moves vertically, that is, it moves automatically towards the target height. If a posture deviation is detected, proceed to step S5. If an obstacle is detected in the direction of movement, proceed to step S6. Then continue to move towards the target height. When it reaches the target height, switch the target height and proceed to step S8. S5, When the wall-climbing robot's posture is greater than the posture threshold 1, the system performs posture correction logic, controls the robot's own rotation, and judges the robot's current posture in real time. When the posture is less than the posture threshold 2, the robot exits the posture correction logic. S6, When moving vertically, if there is an obstacle in the direction of movement, the system enters the vertical obstacle avoidance logic. First, it switches to horizontal movement and simultaneously judges the obstacle information in both the vertical and horizontal directions. If there is no obstacle in the horizontal direction but an obstacle in the vertical direction, it moves horizontally until there is no obstacle in the vertical direction, and then continues to step S4. If the horizontal movement distance exceeds the maximum horizontal movement distance in a single step, it switches the target height and continues to step S4. If there are obstacles in both the horizontal and vertical directions, it switches the target height and continues to step S4. S7. When moving horizontally, if there is an obstacle in the direction of movement, the system enters the horizontal obstacle avoidance logic. First, it switches to vertical movement and simultaneously judges the obstacle information in both the vertical and horizontal directions. If there is no obstacle in the vertical direction but an obstacle in the horizontal direction, it moves vertically until there is no obstacle in the horizontal direction, and then continues to step S8. If the vertical movement distance reaches the target height, it switches the target height and continues to step S8. If there are obstacles in both the horizontal and vertical directions, it switches the target height and continues to step S8. S8, robot horizontal movement state. If posture deviation is detected, proceed to step S5; if there is an obstacle during horizontal movement, proceed to step S7; if the robot moves to the maximum distance of a single horizontal movement, proceed to step S4 and enter the vertical movement state. S9. After the robot moves a total horizontal distance to the set width of the tank wall, the robot completes the full-coverage path planning and obstacle avoidance operation. The program exits the full-coverage path planning system, the progress independent control module is activated, the operation motor is turned off, and the robot is remotely controlled by the remote control module to perform the robot retrieval operation.

2. The method for full-coverage path planning and obstacle avoidance of a wall-climbing robot based on lidar as described in claim 1, characterized in that, The host computer is equipped with a lidar point cloud processing module, a full-coverage walking obstacle avoidance module, and an independent control module; The lidar point cloud processing module is used to acquire the posture and relative position of the wall-climbing robot, acquire information about obstacles around the robot, and provide the data to the full-coverage walking obstacle avoidance module to realize its function. The full-coverage walking obstacle avoidance module is used to comprehensively process information based on the instructions of the remote control module and the relevant information provided by the lidar point cloud processing module, and then control the chassis movement through the chassis control module to realize the robot's full-coverage walking obstacle avoidance function on the tank wall. The independent control module is used to receive instructions from the remote control module to control the movement of the wall-climbing robot and the rotation of its working motors. Then, it informs the chassis control module through a communication protocol to realize the remote control operation of the wall-climbing robot.

3. The method for full-coverage path planning and obstacle avoidance of a wall-climbing robot based on lidar as described in claim 1, characterized in that, The wall-climbing robot has two states: vertical movement and horizontal movement. The maximum distance of a single vertical movement is the set map height, and the maximum distance of a single horizontal movement is the movement distance set according to the robot's single working range.

4. The method for full-coverage path planning and obstacle avoidance of a wall-climbing robot based on lidar according to claim 1, characterized in that, In step S4, the robot is constantly moving towards the target height during vertical movement. There are two target heights: one at the highest point of the map and one at the lowest point of the map. The highest point is the initial height at which the robot starts navigation in the upper right corner of the map, and the lowest point is the initial height minus the set map height. After the robot reaches a target height, the target height is switched, which is manifested by the robot moving up and down back and forth. Horizontal leftward movement is inserted between the upward and downward movements to achieve full-coverage path walking operation.

5. The method for full-coverage path planning and obstacle avoidance of a wall-climbing robot based on lidar according to claim 1, characterized in that, In step S5, there are two attitude thresholds: a large attitude threshold 1, which is used to determine whether the robot needs to correct its attitude to avoid excessive deviation from the work trajectory; and a small attitude threshold 2, which is used to correct the robot's attitude according to the smaller threshold when the robot is correcting its attitude, to ensure that the robot is corrected as accurately as possible.

Images