Autonomous control system for mine pipe gripper and its control method

Abstract

The present invention provides a self-control system for a mine pipe gripper and its control method, which relates to the technical field of mine pipe grippers. The system controls the pipe gripping process by means of vision, perception, and planning. The vision recognition part uses a mine visible light and depth camera for composite image acquisition to improve the accuracy of the coordinates of the pipe to be installed. The attitude perception part is responsible for spatially positioning the attitude of the pipe gripper. Finally, the hydraulic arm motion trajectory is planned by calculating the spatial coordinates of the pipe center and the spatial coordinates of the hydraulic arm center, and the inverse solution of the spatial coordinate system is carried out according to the DH algorithm, which is decomposed into the displacement and angle transformation amounts of the telescopic oil cylinder and the slewing motor, so as to achieve the purpose of the pipe gripper autonomously and accurately gripping the pipe; greatly solving the problems of manual operation of the existing mine pipe grippers, lack of intelligent perception, poor operation quality, and large potential safety hazards.

Description

Technical Field

[0001] The present invention relates to the technical field of mine pipe grabbers, and particularly relates to an autonomous control system for a mine pipe grabber and a control method thereof. Background Art

[0002] In coal production, it is necessary to lay conveying pipelines such as compressed air, water supply, and drainage in various roadways above and below the coal mine. However, manual operations need to ensure the safety of underground workers and the production safety at the site.

[0003] At present, more than 95% of the conveying pipelines underground use steel pipes with a diameter of Φ120 - 250, and the weight of a single pipe reaches 0.5t. The laying of underground pipelines mainly uses manual handling and installation. With the application of pipe grabbers in recent years, there is a trend to replace manual labor. Mechanized operations have partially improved work efficiency, but there are still the following obvious problems in its control system:

[0004] 1) Low intelligence

[0005] It is necessary to manually operate the hydraulic valve or remote control at close range, with cumbersome operation. The operators are prone to fatigue, resulting in misoperations.

[0006] 2) Low safety

[0007] There are operation blind spots. The operator needs to constantly change positions, and once there is an operation error or the environment is harsh, it poses a danger to the personal safety of the auxiliary personnel, with extremely high potential safety hazards.

[0008] 3) Low efficiency

[0009] During operation, 1 operator and 4 auxiliary personnel are required, and they need to closely coordinate and connect. Each step needs to be repeatedly confirmed, seriously affecting the pipeline laying efficiency and causing waste of manpower. Summary of the Invention

[0010] In view of the deficiencies of the prior art, the present invention proposes an autonomous control system for a mine pipe grabber, including: a mine pipe grabber, a server, a controller, and a visible light camera, a depth camera, a magnetostrictive sensor, an absolute encoder, a double inclination sensor, and a servo solenoid valve installed on the mine pipe grabber; the visible light camera and the depth camera are wirelessly connected to the server, the magnetostrictive sensor, the absolute encoder, the double inclination sensor, and the servo solenoid valve are respectively electrically connected to the controller, and the controller is wirelessly connected to the server;

[0011] The visible light camera is used to collect RGB images of the scene where the pipeline is located;

[0012] The depth camera is used to collect point cloud data of the on-site environment where the pipeline to be installed is located;

[0013] The magnetostrictive sensor is used to collect the stroke change amount of the telescopic oil cylinder;

[0014] The absolute encoder is used to collect the rotation angle of the slewing mechanism;

[0015] The dual inclination sensors are used to collect the inclination angle of the pipe gripper body;

[0016] The server is used to plan the trajectory of the hydraulic arm movement according to the RGB image and the point cloud data, decompose the inverse solution of the space coordinate system to obtain the displacement and angle transformation amounts of the telescopic cylinder and the slewing motor, and transmit them to the controller, and control the opening and closing degree of the servo solenoid valve through the controller to control the telescopic cylinder and the slewing motor.

[0017] A method for autonomous control of a mine pipe gripper, which is implemented based on the described autonomous control system of the mine pipe gripper, and the method includes:

[0018] Step 1: Construct a cylindrical point cloud segmentation network model, collect the point cloud information of the pipeline to be installed, fit the pipeline cylinder surface and obtain the axis vector;

[0019] The said Step 1 includes:

[0020] Step 1.1: Construct a cylindrical point cloud segmentation model and perform model training based on the PointNet++ framework; including:

[0021] Step 1.1.1: According to the point cloud data set containing cylinders and the collected pipeline point cloud data, construct a point cloud data set for cylindrical surface segmentation;

[0022] Step 1.1.2: Denoise and enhance the point cloud data set, and standardize the point cloud data volume through resampling;

[0023] Step 1.1.3: Divide the point cloud data after standardization processing into training set, validation set and test set;

[0024] Step 1.1.4: Determine the training-related parameters based on the PointNet++ framework, train the point cloud data set until the error converges, and obtain the cylindrical surface segmentation network model.

[0025] Step 1.2: Use a depth camera to collect the point cloud data of the on-site environment where the pipeline to be installed is located, and input it into the point cloud segmentation model obtained in Step 1.1 after noise reduction and standardized resampling to realize the segmentation and screening of the pipeline cylindrical point cloud;

[0026] Step 1.3: Use the spherical fitting MSAC algorithm to fit the cylindrical surface of the pipeline cylindrical point cloud;

[0027] Step 1.4: Obtain the axis direction according to the fitted pipeline cylinder surface.

[0028] Step 2: Collect the RGB image of the pipeline and locate the pipeline centroid in combination with the axis position; including:

[0029] Step 2.1: Use a pre-calibrated visible light camera to collect the RGB image of the scene where the pipeline is located;

[0030] Step 2.2: Perform gray-scale and binary processing on the collected RGB image, and obtain the position of the pipeline center mark through morphological methods;

[0031] Step 2.3: Use the pre-calibrated external parameters of the camera to calculate the pipeline cylinder points corresponding to the pixels in the pipeline center mark area according to the pipeline cylinder surface fitted in Step 1.3;

[0032] Step 2.4: Calculate the normal vectors of the cylinder points in the pipeline center mark area respectively, take the average value of the normal vectors as the radial vector at the centroid, and the intersection point of the radial vector and the pipeline cylinder surface is the pipeline centroid point.

[0033] Step 3: Calculate the spatial coordinates of the center of the hydraulic arm by collecting the attitude information of the pipe gripper; including:

[0034] Use a dual-inclination sensor to collect the inclination angle of the pipe gripper body in real time, and convert the body inclination angle into the offset angle of the hydraulic arm relative to the camera world coordinate system;

[0035] Use an absolute encoder to collect the rotation angle of the pipe gripper slewing mechanism, and convert the rotation angle of the slewing mechanism into the transformation radian of the front end effector of the hydraulic arm in the vertical and horizontal planes;

[0036] Use a hysteresis sensor to collect the stroke change of the telescopic cylinder, and calculate the spatial coordinates of the center of the hydraulic arm in combination with the stroke change of the telescopic cylinder; specifically expressed as: use the geometric parameters of the manipulator body to establish the conversion relationship between the change of the telescopic cylinder and the joint angle of the hydraulic manipulator to realize the conversion from the driving space to the joint space; collect the joint angles of the manipulator through the encoders installed on each joint of the manipulator; use the fixed angle method to solve the inverse kinematics of the manipulator to obtain the relationship between the joint angles of the manipulator and the spatial coordinates of the center of the manipulator, and realize the conversion from the joint space to the operation space of the manipulator.

[0037] Step 4: Plan the motion trajectory of the hydraulic arm according to the pipeline centroid coordinates and the spatial coordinates of the center of the hydraulic arm to realize automatic pipeline grasping; including:

[0038] Step 4.1: Through the inverse solution of the space coordinate system, decompose and obtain the displacement and angle transformation amounts of the telescopic cylinder and the slewing motor;

[0039] Step 4.2: Generate a collision-free trajectory from the starting point to the end point, and expand and search in the n-dimensional space to generate a search trajectory, where n represents the degree of freedom of the manipulator;

[0040] Step 4.3: According to the inverse kinematics of the robotic arm, track the searched trajectory in real time, and realize automatic grasping of the pipeline by controlling the movement of the robotic arm.

[0041] The beneficial effects of the present invention are as follows:

[0042] The present invention proposes an autonomous control system and its control method for a mine pipe gripper, which overcomes the deficiencies of the traditional pipe gripper control system, such as poor safety, low autonomy, and waste of labor. Especially for the situation of harsh on-site environment and shortage of personnel, it can autonomously perform pipeline installation operations without manual assistance, truly realizing the autonomy and efficiency of the mine pipe gripper. The whole process requires no manual intervention, and its accuracy and efficiency are much higher than that of manual work, and it truly meets the urgent need of the coal mine for intelligentization. Brief Description of the Drawings

[0043] Figure 1 It is a schematic diagram of the autonomous control system of the mine pipe gripper in the present invention;

[0044] Figure 2 It is a schematic diagram of the principle of the autonomous control method of the mine pipe gripper in the present invention;

[0045] In the figure, 1. Telescopic oil cylinder; 2. Rotary motor; 3. Visible light camera; 4. Depth camera; 5. Magnetostrictive sensor; 6. Absolute encoder; 7. Dual inclination sensor; 8. Servo solenoid valve. Detailed Embodiments

[0046] The following further describes the invention in conjunction with the drawings and specific implementation examples.

[0047] An autonomous control system for a mine pipe gripper includes: a mine pipe gripper, a server, a controller (PLC), and an RGB visible light camera 3, a depth camera 4, a magnetostrictive sensor 5, an absolute encoder 6, a dual inclination sensor 7, and a servo solenoid valve 8 installed on the mine pipe gripper; the RGB visible light camera 3 and the depth camera 4 are wirelessly connected to the server, and the magnetostrictive sensor 5, the absolute encoder 6, the dual inclination sensor 7, and the servo solenoid valve 8 are respectively electrically connected to the controller, and the controller is wirelessly connected to the server; as Figure 1 shown, the magnetostrictive sensor 5 is installed in the telescopic oil cylinder 1, the absolute encoder 6 and the dual inclination sensor 7 are installed on the hydraulic arm of the mine pipe gripper, the visible light camera 3 and the depth camera 4 are installed on the rotary motor 2, and the servo solenoid valve 8 is installed on the body of the mine pipe gripper.

[0048] The visible light camera is used to collect RGB images of the scene where the pipeline is located;

[0049] The depth camera is used to collect point cloud data of the on-site environment where the pipeline to be installed is located;

[0050] The magnetostrictive sensor is used to collect the stroke change of the telescopic oil cylinder;

[0051] The absolute encoder is used to collect the rotation angle of the slewing mechanism;

[0052] The dual inclination sensor is used to collect the inclination angle of the pipe gripper body;

[0053] The server is used to perform trajectory planning for the movement of the hydraulic arm based on the RGB image and point cloud data, perform inverse solution of the spatial coordinate system according to the DH algorithm (Diffie–Hellman key exchange, abbreviated as D-H), decompose it into the displacement and angle transformation of the telescopic oil cylinder and slewing motor, and transmit it to the controller. The controller controls the opening and closing degree of the servo solenoid valve to control the telescopic oil cylinder and slewing motor.

[0054] As Figure 2 shown, the system uses vision, perception, and planning to control the pipe gripping process, realizes full-process autonomous decision-making from pipe identification to gripping, and greatly solves the problems of manual operation of existing mine pipe grippers, lack of intelligent perception, poor operation quality, and large safety hazards.

[0055] The visual recognition part uses a mine visible light and depth camera to collect composite images to improve the accuracy of the coordinates of the pipe to be installed. Since there are many cylindrical equipment underground, it is easy to affect visual acquisition. Therefore, a mine visible light camera is used to take pictures of the flange of the pipe as the recognition feature. The pictures are processed by image decoding and noise reduction to obtain the outer contour material of the flange, which is fitted and compared with the pre-learned sample library. After confirming that the object to be measured is a pipe. Then, the dot matrix collected by the mine depth camera is subjected to feature deep learning. Since the distance may cause the edge of the collected dot matrix to be not smooth, after superimposing the results of the deep learning with the outer contour material of the flange, the spatial coordinates of the pipe center in the camera world coordinate system are accurately calculated.

[0056] The attitude perception part is mainly responsible for spatially positioning the attitude of the pipe gripper. The mine dual inclination sensor is used to collect the inclination angle of the fuselage in real time, the mine absolute encoder is responsible for collecting the rotation angle of the slewing mechanism, and the mine magnetic hysteresis sensor is responsible for collecting the stroke change of the telescopic oil cylinder. The system calculates the offset angle of the hydraulic arm relative to the camera world coordinate system according to the inclination angle of the fuselage, calculates the transformation radian of the front end gripper of the hydraulic arm in the vertical and horizontal planes according to the rotation angle of the slewing mechanism, and then combines the stroke change of the telescopic oil cylinder for composite conversion to accurately calculate the spatial coordinates of the hydraulic arm center in real time.

[0057] The path planning and execution part mainly controls the movement trajectory of the hydraulic arm and the solenoid valve. The path planning algorithm plans the movement trajectory of the hydraulic arm by calculating the spatial coordinates of the pipeline center and the hydraulic arm center, performs inverse solution of the spatial coordinate system according to the DH algorithm, decomposes it into the displacement and angle transformation amounts of the telescopic oil cylinder and the slewing motor, and combines the dead zone and inertia of the hydraulic system to accurately control the opening degree of the mining servo solenoid valve reasonably, and finally realizes the purpose of the pipe gripper autonomously and accurately grasping the pipeline.

[0058] Based on the above implementation principle, a method for autonomous control of a mining pipe gripper provided by the present invention is realized based on the above-mentioned autonomous control system for a mining pipe gripper, and the method includes:

[0059] Step 1: Construct a cylindrical point cloud segmentation network model, collect the point cloud information of the pipeline to be installed, fit the pipeline cylinder surface and obtain the axis vector; specifically expressed as:

[0060] Step 1.1: Construct a cylindrical point cloud segmentation model and train the cylindrical point cloud segmentation model based on the PointNet++ framework; specifically expressed as:

[0061] Step 1.1.1: Use the public point cloud data set containing cylinders and the pipeline point cloud data collected on site to construct a point cloud data set for cylindrical surface segmentation;

[0062] Step 1.1.2: Denoise and enhance the point cloud data set, and then standardize the point cloud data volume through resampling;

[0063] Step 1.1.3: Divide the point cloud data after the standardization process into a training set, a validation set and a test set to complete the preprocessing of the data set;

[0064] Step 1.1.4: Build a network training environment, determine the training-related parameters based on the PointNet++ framework, train the point cloud data set until the error converges, obtain the cylindrical surface segmentation network model and complete the verification;

[0065] Step 1.2: Use a depth camera to collect the point cloud data of the on-site environment where the pipeline to be installed is located, and input it into the point cloud segmentation model obtained in Step 1.1 after noise reduction and standardized resampling, so as to realize the segmentation and screening of the pipeline cylindrical point cloud;

[0066] Step 1.3: Use the spherical fitting MSAC algorithm to fit the cylindrical surface of the pipeline cylindrical point cloud;

[0067] Step 1.4: Directly obtain the axis direction according to the fitted pipeline cylinder surface;

[0068] Step 2: Collect the RGB image of the pipeline and combine the axis position to realize the centroid positioning of the pipeline; specifically expressed as:

[0069] Step 2.1: Use a pre-calibrated visible light camera (whose external parameters have been calibrated with the depth camera) to collect the RGB image of the scene where the pipeline is located;

[0070] Step 2.2: Perform grayscale and binarization processing on the collected RGB image, and obtain the position of the pipeline center mark through morphological methods;

[0071] Step 2.3: Using the pre-calibrated external parameters, calculate the pipeline cylinder points corresponding to the pixels in the pipeline center mark area based on the pipeline cylinder surface fitted in Step 1.3;

[0072] Step 2.4: Calculate the normal vectors of the cylinder points in the above-mentioned mark area respectively, regard the average value of the normal vectors as the radial vector at the centroid, and the intersection point of the radial vector and the pipeline cylinder surface is the pipeline centroid point, thus completing the determination of the pipeline centroid position. So far, the attitude of the pipeline and the spatial position of the centroid have been all calculated.

[0073] Step 3: Calculate the spatial coordinates of the center of the hydraulic arm by collecting the attitude information of the pipe gripper; specifically described as:

[0074] Use a mining double-inclination sensor to collect the inclination angle of the pipe gripper body in real time, and convert the body inclination angle into the offset angle of the hydraulic arm relative to the camera world coordinate system through the homogeneous transformation matrix of the robotic arm;

[0075] Use a mining absolute encoder to collect the rotation angle of the slewing mechanism, and convert the rotation angle of the slewing mechanism into the transformation radian of the front end effector of the hydraulic arm in the vertical and horizontal planes through the homogeneous transformation matrix of the robotic arm;

[0076] Use a mining hysteresis sensor to collect the stroke change of the telescopic cylinder, and perform composite conversion in combination with the stroke change of the telescopic cylinder to accurately calculate the spatial coordinates of the center of the hydraulic arm in real time; specifically described as:

[0077] The controller outputs a PWM signal through an internal proportional amplifier board to control the opening of the spool of the proportional solenoid valve, thereby controlling the flow rate and pressure of the hydraulic oil in the cylinder to achieve precise control of the cylinder stroke. Use the geometric parameters of the robotic arm body to establish the conversion relationship between the change of the telescopic cylinder and the joint angle of the hydraulic robotic arm, and realize the conversion from the driving space to the joint space;

[0078] Collect the joint angles of the robotic arm through the encoders installed on each joint of the robotic arm to realize the closed-loop feedback of the joint angles of each joint of the robotic arm and accurately control the joint angles of the robotic arm;

[0079] Use the fixed angle method to solve the inverse kinematics of the robotic arm, and obtain the relationship between the joint angles of the hydraulic robotic arm and the spatial coordinates of the center of the hydraulic robotic arm, realizing the conversion from the joint space to the operation space of the robotic arm.

[0080] Step 4: Plan the movement trajectory of the hydraulic arm according to the spatial coordinates of the pipeline center and the spatial coordinates of the hydraulic arm center to achieve the purpose of the pipe gripper autonomously and accurately grasping the pipeline; specifically expressed as:

[0081] Step 4.1: Perform inverse solution of the spatial coordinate system according to the Diffie-Hellman algorithm, and decompose it into the displacement and angle transformation amounts of the telescopic cylinder and the slewing motor;

[0082] Step 4.2: Use the rapidly exploring random tree algorithm to generate a collision-free trajectory from the starting point to the ending point. Assuming that the robotic arm has n degrees of freedom, complete its expansion and search in the n-dimensional space to generate the search trajectory;

[0083] Step 4.3: Use the inverse kinematics of the robotic arm to perform real-time tracking of the searched trajectory, and use the conversion relationship in different spaces in Step 3 to achieve the motion control of the robotic arm.

[0084] An autonomous control system for a mine pipe gripper proposed by the present invention, through its three parts of visual recognition and analysis, attitude perception and positioning, and hydraulic arm path planning and control, can enable the pipe gripper to autonomously identify the pipeline to be installed, and autonomously perform pipe grasping operations and installation operations, fully achieving the purpose of unmanned operation in the whole process of pipe grasping work.

[0085] This autonomous control system can reduce the labor intensity of workers, improve work efficiency, greatly avoid problems such as low cooperation efficiency, large visual blind areas, and cumbersome operation steps during manual operation, and reduce the occurrence of safety accidents at the construction site.

Claims

1. An autonomous control system for a mine pipe grabber, characterized in that Including: A mine pipe grabber, a server, a controller, a visible light camera, a depth camera, a magnetostrictive sensor, an absolute encoder, a dual inclination sensor, and a servo solenoid valve installed on the mine pipe grabber; the visible light camera and the depth camera are wirelessly connected to the server, the magnetostrictive sensor, the absolute encoder, the dual inclination sensor, and the servo solenoid valve are electrically connected to the controller respectively, and the controller is wirelessly connected to the server; The visible light camera is used to collect RGB images of the scene where the pipeline is located; The depth camera is used to collect point cloud data of the on-site environment where the pipeline to be installed is located; The magnetostrictive sensor is used to collect the stroke change of the telescopic oil cylinder; The absolute encoder is used to collect the rotation angle of the slewing mechanism; The dual inclination sensor is used to collect the inclination angle of the body of the pipe grabber; The server is used to perform trajectory planning for the movement of the hydraulic arm according to the RGB image and the point cloud data, decompose and obtain the displacement and angle transformation of the telescopic oil cylinder and the slewing motor through inverse kinematics in the space coordinate system, and transmit them to the controller, and control the opening and closing degree of the servo solenoid valve through the controller to control the telescopic oil cylinder and the slewing motor; A mine pipe grabber autonomous control system realizes automatic pipe grabbing by the following method, including: Step 1: Construct a cylindrical point cloud segmentation network model, collect the point cloud information of the pipeline to be installed, fit the cylindrical surface of the pipeline and obtain the axis vector; Step 2: Collect the RGB image of the pipeline and locate the centroid of the pipeline in combination with the axis position; Step 2.1: Use a pre-calibrated visible light camera to collect the RGB image of the scene where the pipeline is located; Step 2.2: Perform gray-scale and binary processing on the collected RGB image, and obtain the position of the pipeline center mark through morphological methods; Step 2.3: Utilize the pre-calibrated external parameters of the camera, and calculate the cylindrical points of the pipeline corresponding to the pixels in the pipeline center mark area respectively according to the cylindrical surface of the pipeline fitted in Step 1.3; Step 2.4: Calculate the normal vectors of the cylindrical points in the pipeline center mark area respectively, take the average value of the normal vectors as the radial vector at the centroid, and the intersection point of the radial vector and the cylindrical surface of the pipeline is the centroid point of the pipeline; Step 3: Calculate the spatial coordinates of the center of the hydraulic arm by collecting the attitude information of the pipe grabber; Use the dual inclination sensor to collect the inclination angle of the body of the pipe grabber in real time, and convert the body inclination angle into the offset angle of the hydraulic arm relative to the camera world coordinate system; Use the absolute encoder to collect the rotation angle of the slewing mechanism of the pipe grabber, and convert the rotation angle of the slewing mechanism into the transformation radian of the front end effector of the hydraulic arm in the vertical and horizontal planes; Use the magnetic sensor to collect the stroke change of the telescopic oil cylinder, and calculate the spatial coordinates of the center of the hydraulic arm in combination with the stroke change of the telescopic oil cylinder; Step 4: Perform trajectory planning for the movement of the hydraulic arm according to the centroid coordinates of the pipeline and the spatial coordinates of the center of the hydraulic arm to realize automatic pipe grabbing.

2. The autonomous control system of a mine pipe grabber according to claim 1, characterized in that, The said Step 1 includes: Step 1.1: Construct a cylindrical point cloud segmentation model and perform model training based on the PointNet++ framework; Step 1.2: Use a depth camera to collect the point cloud data of the on-site environment where the pipeline to be installed is located. After noise reduction and standardized resampling, input it into the point cloud segmentation model obtained in Step 1.1 to achieve the segmentation and screening of the pipeline column point cloud; Step 1.3: Use the spherical fitting algorithm to perform cylindrical surface fitting on the pipeline column point cloud; Step 1.4: Obtain the axis direction according to the fitted pipeline column surface.

3. The autonomous control system of a mine pipe grabber according to claim 2, characterized in that, The said Step 1.1 includes: Step 1.1.1: According to the point cloud data set containing cylinders and the collected pipeline point cloud data, construct a point cloud data set for column surface segmentation; Step 1.1.2: Denoise and enhance the point cloud data set, and standardize the amount of point cloud data through resampling; Step 1.1.3: Divide the point cloud data after standardized processing into a training set, a validation set and a test set; Step 1.1.4: Based on the PointNet++ framework, determine the training-related parameters, and train the point cloud data set until the error converges to obtain a cylindrical surface segmentation network model.

4. The autonomous control system of a mine pipe grabber according to claim 1, characterized in that, The stroke change of the telescopic oil cylinder is collected by using a hysteresis sensor, and the central space coordinates of the hydraulic arm are calculated in combination with the stroke change of the telescopic oil cylinder. The specific description is as follows: Using the geometric parameters of the robotic arm body, establish the conversion relationship between the change of the telescopic oil cylinder and the joint angle of the hydraulic robotic arm to achieve the conversion from the driving space to the joint space; The joint angles of the robotic arm are collected by the encoders installed on each joint of the robotic arm; Using the fixed angle method, solve the inverse kinematics of the robotic arm to obtain the relationship between the joint angles of the robotic arm and the central space coordinates of the robotic arm, and achieve the conversion from the joint space to the working space of the robotic arm.

5. The autonomous control system of a mine pipe grabber according to claim 1, characterized in that, The said Step 4 includes: Step 4.1: Through the inverse solution of the space coordinate system, decompose to obtain the displacement and angle transformation amounts of the telescopic oil cylinder and the slewing motor; Step 4.2: Generate a collision-free trajectory from the starting point to the ending point, and expand and search in the n-dimensional space to generate a search trajectory, where n represents the degree of freedom of the robotic arm; Step 4.3: According to the inverse kinematics of the robotic arm, perform real-time tracking on the searched trajectory, and realize automatic grasping of the pipeline by controlling the movement of the robotic arm.

Images