System and method for controlling pipeline polishing robot

Abstract

The invention discloses a system and method for controlling a pipeline polishing robot, and relates to the technical field of robot control, the system is composed of a plurality of functional modules, and the system comprises a reducing self-adaptive preprocessing module which obtains a pipe diameter curve through laser continuous scanning, splices continuously scanned point cloud data in real time through an SLAM algorithm, and obtains a pipe diameter curve; constructing a dynamic three-dimensional topological model of the pipeline; attitude data of the robot are obtained, and the offset of the robot in the central axis of the pipeline is calculated; obtaining the monitoring environment temperature in the pipeline, constructing a dynamic thermal expansion model of the pipeline through thermal conduction finite element analysis, and correcting the pipe diameter value according to the thermal expansion model; according to the offset and the corrected pipe diameter value, the expansion and contraction amount of the umbrella-shaped structure is calculated through a fuzzy PID controller, and an umbrella-shaped structure expansion and contraction instruction is generated; the dust collection head is synchronously unfolded to the pipe wall; and a multi-process cooperative control module.

Description

Technical Field

[0001] The present invention relates to the technical field of robot control, and specifically to a system and method for controlling a pipeline grinding robot. Background Art

[0002] As a key water conveyance structure of a water conservancy project, the structure system of a water conveyance tunnel consists of three main parts: an intake sluice, a main tunnel section, and an end control chamber, forming a complete water flow conveyance channel; the intake sluice undertakes the functions of water level regulation and flow control, adopts a reinforced concrete frame structure and is equipped with a two-way opening and closing gate; the main tunnel section is the core water conveyance carrier of the tunnel, adopts a lining structure of Q345B low-alloy steel plate with a wall thickness of 12 mm, and the pipe diameter ranges from DN1200 to DN1300; the end gate chamber section is provided with an arc-shaped steel working gate, and is equipped with a hydraulic hoisting system to achieve precise flow regulation; during the continuous operation of this structure system, due to the superposition of multiple factors such as hydraulic load and environmental erosion, serious deterioration of material properties has occurred.

[0003] Traditional anti-corrosion repair mostly relies on manual operation inside the pipeline. The tunnel inlet is located at the bottom of the reservoir. Construction workers can only enter and exit through a single-side exit, and the ventilation conditions are extremely poor. Limited-space operations are prone to accidents such as asphyxiation and poisoning; the manual rust removal speed is only 0.5 - 1 m² / h, and due to pipe diameter limitations, it is usually only applicable to manholes with a diameter above DN800. For a 281 m pipeline, it takes more than 30 days; the thickness fluctuation of the manually sprayed coating is >15%, especially in the variable-diameter section, there are prone to missed coating and accumulation, resulting in local anti-corrosion failure; commercially available robots are mostly designed for small pipe diameters below DN600, and can only perform local anti-corrosion of welds, unable to meet the requirements of continuous operation for large-diameter (>DN1000) pipelines, and traditional equipment lacks the ability to adapt to pipe diameter changes. In the variable-diameter area, manual intervention is required to adjust the mechanical structure, interrupting the continuity of construction. Summary of the Invention

[0004] To achieve the above objectives, the present invention is realized through the following technical solutions: A system for controlling a pipeline grinding robot, comprising: A variable-diameter adaptive preprocessing module, which obtains the pipe diameter curve through continuous laser scanning, and real-time splices the continuously scanned point cloud data through the SLAM algorithm to construct a dynamic three-dimensional topological model of the pipeline; obtains the attitude data of the robot, calculates the offset of the robot on the central axis inside the pipeline; obtains the monitored environmental temperature inside the pipeline, constructs a dynamic thermal expansion model of the pipeline through finite element analysis of heat conduction, and corrects the pipe diameter value according to the thermal expansion model; according to the offset and the corrected pipe diameter value, calculates the telescopic amount of the umbrella-shaped structure through a fuzzy PID controller, and generates an instruction for the telescopic amount of the umbrella-shaped structure; The multi - process collaborative control module dynamically sets the spraying reference speed and relationship according to the pipe diameter curve and coating thickness requirements; it uses the YOLOv5 model to detect foreign objects on the pipe wall in real time. When the height of the identified obstacle is greater than the height threshold c, it pauses the movement, triggers cleaning, and then continues the process; if it still cannot be cleared, it marks the position and notifies the console for manual intervention. When the height of the identified obstacle is less than or equal to the height threshold c, it pauses the movement and triggers local supplementary grinding. The coating thickness closed - loop control module obtains the thickness feedback in real time, compares it with the target value, adjusts the rotation speed of the spraying motor, and performs floating compensation on the nozzle.

[0005] Furthermore, the process of obtaining the pipe diameter curve is as follows: Perform circular arc fitting on the spliced point cloud data, calculate the inner diameter value of each cross - section; identify the variable diameter area through curvature calculation, and generate a segmented and continuous pipe diameter curve.

[0006] Furthermore, the process of real - time splicing through the SLAM algorithm is as follows: Apply the LIO - SLAM algorithm and achieve dynamic splicing through the following steps: S101: Use IMU data, accelerometers and gyroscopes to calculate the pose change of the robot in real time; S102: Align the point cloud data of adjacent frames through the ICP algorithm; S103: Based on the SLAM framework, construct a dynamic three - dimensional topological model of the pipeline from the spliced point cloud data; Use voxel filtering to reduce the point cloud density while retaining geometric features; remove noise points through statistical outlier detection.

[0007] Furthermore, the process of constructing the dynamic three - dimensional topological model of the pipeline is as follows: Based on the dense point cloud data, cover the entire pipeline cross - section and generate the inner wall of the pipeline with a complex shape; each frame of point cloud data contains spatial point coordinate information, and filter the point cloud to remove isolated points; Use the ICP algorithm to match a newly obtained frame of point cloud with the previously constructed map, find the rigid body transformation between the two frames of point cloud, and form a global map by integrating the results of multiple local registrations; based on the registered point cloud data, use computer vision algorithms to extract the key geometric features inside the pipeline and update them dynamically.

[0008] Furthermore, the process of calculating the offset of the robot's central axis in the pipeline is as follows: The angular velocity and acceleration of the robot body are obtained through sensors. Based on the three-dimensional topological model of the pipeline constructed by laser scanning, the central axis of the pipeline is extracted as the ideal trajectory. Combining the attitude angle data of the IMU and the robot kinematic model, the actual central axis position of the robot is calculated. The attitude angles measured by the IMU are converted into the pose in the robot body coordinate system, and the translational displacement of the robot is obtained through the odometer. The actual central axis position is compared with the ideal trajectory to calculate the deviation.

[0009] Further, the process of correcting the pipe diameter value is as follows: Based on the pipeline material properties, the model is meshed, the continuous pipeline structure is discretized into many small cells, and appropriate boundary conditions are set according to the actual situation. The finite element method is used to solve the heat conduction equation to predict the temperature field distribution of each part of the pipeline changing with time under given conditions. According to the temperature field distribution, the thermal expansion of each position of the pipeline is calculated section by section. The average value of the inner and outer wall temperatures is used as the correction basis.

[0010] Further, the specific process of generating the telescopic instruction for the umbrella structure is as follows: The three parameters of the traditional PID, the proportional gain Kp, the integral gain Ki, and the derivative gain Kd, are dynamically adjusted through fuzzy logic. Its input is the error and the error change rate, and the output is the PID parameter correction value. Mamdani fuzzy inference is used to calculate the weight of each rule. Through the centroid method, the fuzzy output is converted into accurate Kp, Ki, and Kd values. According to the corrected Kp, Ki, and Kd values, the telescopic amount is adjusted in real time.

[0011] Further, the process of dynamically setting the spraying reference speed and relationship is as follows: Through the determination of the spraying reference speed, based on the principle of conservation of paint flow, precise control is achieved for the dynamic balance of the pipe diameter, coating thickness, and paint flow. The real-time pipe diameter is obtained through laser scanning. Combining the temperature sensor data, the thermal expansion of the pipeline is calculated to obtain the corrected actual pipe diameter. The effective flow is adjusted according to the ambient temperature to offset the change in paint fluidity caused by high temperature. The spraying speed is set according to the output efficiency of the spraying equipment under different conditions.

[0012] Further, the process of the nozzle performing floating compensation is as follows: The electromagnetic eddy current is used to automatically detect the coating thickness. The operator sets the target coating thickness through the HMI. The motor speed is dynamically adjusted according to the deviation size to eliminate the long-term deviation. The force sensor and displacement sensor are integrated to detect the contact force and relative position between the nozzle and the workpiece in real time. The floating resistance of the nozzle is automatically adjusted according to the contact force, and the nozzle angle is adjusted through floating compensation.

[0013] A method for controlling a pipeline grinding robot, comprising the following steps: Step 1: Continuously scan the pipe diameter curve through a laser, and splice the continuously scanned point cloud data in real time through the SLAM algorithm to construct a dynamic three-dimensional topological model of the pipeline; obtain the attitude data of the robot, calculate the offset of the robot on the central axis inside the pipeline; obtain the monitored environmental temperature inside the pipeline, construct a dynamic thermal expansion model of the pipeline through finite element analysis of heat conduction, and correct the pipe diameter value according to the thermal expansion model; according to the offset and the corrected pipe diameter value, calculate the telescopic amount of the umbrella structure through a fuzzy PID controller, and generate an instruction for the telescopic amount of the umbrella structure; Step 2: Dynamically set the spraying reference speed and relationship according to the pipe diameter curve and the coating thickness requirement; detect foreign objects on the pipe wall in real time through the YOLOv5 model. When the height of the identified obstacle is greater than the height threshold c, pause the progress, trigger the clearance, and continue the process; if the clearance still cannot be achieved, mark the position and notify the console for manual intervention; When the height of the identified obstacle is less than or equal to the height threshold c, pause the progress and trigger local supplementary grinding; Step 3: Obtain the thickness feedback in real time, compare it with the target value, adjust the rotation speed of the spraying motor, and perform floating compensation on the nozzle.

[0014] A system and method for controlling a pipeline grinding robot provided by the present invention have the following beneficial effects: (1) The present invention dynamically adjusts the telescopic amount of the umbrella structure through a fuzzy PID controller, combines the real-time offset and the corrected pipe diameter value, and realizes dynamic compensation with millimeter-level accuracy; the fuzzy PID controller dynamically adjusts the Kp / Ki / Kd parameters through fuzzy logic, suppresses sensor noise and external interference, and the system can still operate stably in harsh environments such as dust and vibration.

[0015] (2) The present invention uses a fuzzy rule base to automatically optimize the PID parameters, eliminating the need for manual repeated tuning, shortening the debugging time, dynamically adjusting the telescopic range of the umbrella structure by correcting the pipe diameter value, supporting multiple pipeline specifications, and the same device can be adapted to pipelines with different diameters without replacing hardware; the closed-loop feedback mechanism combined with the fuzzy PID controller ensures that the nozzle / sensor is always in the best working position. Description of the Drawings

[0016] Figure 1 It is a schematic diagram of the system flow of the present invention; Figure 2 It is a schematic diagram of the overall method of the present invention. Detailed Embodiment

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment 1 Please refer to Figure 1 , Embodiment 1 of the present application provides a system for controlling a pipeline grinding robot, and the system includes: A variable-diameter adaptive preprocessing module, which obtains the pipe diameter curve through continuous laser scanning, splices the continuously scanned point cloud data in real time through the SLAM algorithm, and constructs a dynamic three-dimensional topological model of the pipeline; obtains the attitude data of the robot, calculates the offset of the central axis of the robot in the pipeline; obtains the monitored environmental temperature in the pipeline, constructs a dynamic thermal expansion model of the pipeline through finite element analysis of heat conduction, and corrects the pipe diameter value according to the thermal expansion model; according to the offset and the corrected pipe diameter value, calculates the multi-degree-of-freedom telescopic amount of the umbrella structure through a fuzzy PID controller, and generates an umbrella structure telescopic instruction; the dust suction head is synchronously deployed to the pipe wall; The total length of the robot is 6.8 meters. The equipment consists of a power tractor, a grinding and rust-removing vehicle, a dust suction and cleaning vehicle, a spraying and anti-corrosion vehicle, as well as a wireless remote control module and a control box; Power tractor: It is equipped with a traction motor and a working lithium battery pack inside. Wireless bridge modules, LED lights, system switches, cameras, laser scanners and other sensors are arranged in front of the vehicle; Grinding and rust-removing vehicle: It is the key improved part of the robot. An innovative umbrella-shaped radial grinding structure that can operate continuously is used. The central main shaft supports the entire grinding assembly, enabling it to rotate 360°; Dust suction and cleaning vehicle: It is equipped with a sealed dust collection bin and a telescopic and rotating dust suction head inside; Spraying vehicle: It is equipped with two groups of metering paint pumps and two material bins inside. A high-speed spraying motor is installed at the front end of the metering pump, and a sprayer is fixed at the end of the motor shaft. Paint and curing agent enter the inside of the sprayer through the feeding copper pipe at the same time, are fully mixed through the conical surface of the sprayer, and then are sprayed out at high speed by the sprayer; Wireless module and control box: The wireless module uses a pair of high-power bridges, one is installed on the battery pack tractor, and the other is connected to the control box; there is a display and a touch screen on the control box cover, and a control system is built inside.

[0019] Obtain the pipe diameter curve: Adopt a multi-line lidar with a scanning resolution of 0.1 mm, obtain 100 frames of point cloud data per second, with a ranging range of 30 - 500 mm, install it on the central axis of the robot body, and drive it to rotate 360° for scanning through a servo motor; if the pipeline diameter is large, multiple groups of laser sensors can be deployed to synchronously scan from different angles to avoid occlusion problems; perform circular arc fitting on the spliced point cloud data to calculate the inner diameter value of each cross-section; perform denoising and plane segmentation on the spliced point cloud data to extract the contour points of the pipeline cross-section; use the least squares method or the RANSAC algorithm to perform circular arc fitting on the contour points to solve for the optimal center and radius; obtain the inner diameter value of each cross-section according to the fitting result, cut out multiple cross-section point sets along the pipeline axis from the spliced and corrected three-dimensional point cloud data; perform two-dimensional circular arc fitting on the contour points of each cross-section to obtain the optimal center and radius of the cross-section; take twice the fitting radius as the inner diameter value of the cross-section to form cross-section-by-cross-section pipe diameter distribution data, and perform pipe diameter change trend analysis and variable diameter area identification; Real-time splicing through the SLAM algorithm: Adopt the LIO-SLAM algorithm application and achieve dynamic splicing through the following steps: S101: Utilize IMU data, accelerometers and gyroscopes to real-time calculate the pose change of the robot; S102: Align the point cloud data of adjacent frames through the ICP algorithm to eliminate motion distortion; S103: Based on the SLAM framework, construct the spliced point cloud data into a dynamic three-dimensional topological model of the pipeline; Adopt voxel filtering to reduce the point cloud density while retaining geometric features; remove noise points through statistical outlier detection; Construct a dynamic three-dimensional topological model of the pipeline: Based on the dense point cloud data, covering the entire pipeline cross-section, generate a detailed model of the inner wall of the pipeline with a complex shape; each frame of point cloud data contains the spatial point coordinate information in the lidar coordinate system, reflecting the geometric shape of the pipeline inner wall at a certain moment; due to environmental factors such as dust and water vapor, the original point cloud data may contain noise. Perform filtering on the point cloud to remove isolated points or outliers to improve the accuracy of subsequent processing; Adopt the ICP algorithm to match a newly acquired frame of point cloud with the previously constructed map, find the best rigid body transformation between the two frames of point cloud to make them as coincident as possible, form a consistent global map by integrating the results of multiple local registrations, solve the cumulative error, and improve the quality of the three-dimensional model; Based on the registered point cloud data, computer vision algorithms are used to extract the key geometric features inside the pipeline, such as diameter changes, bending degrees, and branching situations, etc., and then a detailed topological structure is established; as the robot moves forward, new point cloud data is continuously added to the existing map, realizing the dynamic update of the three-dimensional model of the pipeline, ensuring high mapping accuracy even during long-distance exploration; Calculate the offset of the robot's central axis inside the pipeline: By installing a three-axis gyroscope and a three-axis accelerometer, measure the angular velocity and acceleration of the robot body; Gyroscope: Provide angular velocity data, and the attitude angle can be calculated by integration; Accelerometer: Calculate the static attitude angle through the direction of gravity, such as pitch angle and roll angle; Fuse the accelerometer and gyroscope data of the IMU to eliminate noise and improve the attitude estimation accuracy. The formulas for obtaining the pitch angle and roll angle respectively are: Pitch angle: ; Roll angle: ; Where, is the angle of the robot's rotation around the X-axis; , and are the three-axis measurement values of the accelerometer respectively, which are the accelerations from the I inertial measurement unit; is the angle of the robot's rotation around the Y-axis, that is, the roll angle; Extract the central axis of the pipeline as the ideal trajectory through the three-dimensional topological model of the pipeline constructed by laser scanning; combine the attitude angle data of the IMU and the robot kinematic model to calculate the actual central axis position of the robot; convert the attitude angle measured by the IMU into the pose in the robot body coordinate system, and obtain the translational displacement of the robot through the odometer or SLAM algorithm; compare the actual central axis position with the ideal trajectory to calculate the deviation; Calculate the deviation: Detect the distance between the robot and the pipe wall in real time through the laser scanning point cloud data to verify the accuracy of the offset calculation; use the lidar device to scan the surrounding environment to obtain a series of point cloud data describing the surface characteristics of the pipe wall; each point contains the spatial position information relative to the lidar coordinate system; for each or a group of point cloud data representing the pipe wall surface, calculate a normal vector; determine the exact position of the robot itself, that is, the coordinates of the center point in space; use the distance calculation formula to determine the actual distance from the center point of the robot to the pipe wall, and select one or more pipe wall points from the point cloud data as reference points; compare the calculated distance with the expected ideal distance; Taking the offset as the input, the driving motor speed is adjusted through a PID controller to make the robot return to the ideal trajectory; Finite element analysis of heat conduction: Temperature sensors are deployed on the surface of the robot to monitor the temperature change inside the pipeline in real time; Based on the material properties of the pipeline, such as thermal conductivity, specific heat capacity, etc.; and geometric shape, a three-dimensional model of the pipeline is established using professional engineering software; The model is meshed, and the continuous pipeline structure is discretized into many small cells, each cell representing an independent but interconnected small volume; And appropriate boundary conditions are set according to the actual situation, including but not limited to the initial temperature distribution, external heat flux density, thermal impedance between contact surfaces, etc.; The finite element method is used to solve the heat conduction equation to predict the temperature field distribution of each part of the pipeline over time under given conditions; Temperature field distribution: Based on the material properties and geometric shape of the pipeline, a three-dimensional finite element model is constructed and meshed, and the continuous structure is transformed into computable discrete elements; According to the actual working conditions, boundary conditions such as initial temperature, external heat flux, and contact thermal resistance are set to simulate the real heat transfer environment; The heat conduction equation is solved through numerical calculation to obtain the temperature distribution results of each part of the pipeline at different times, forming a dynamic temperature field distribution; Construct a dynamic thermal expansion model of the pipeline: In engineering, the water conveyance tunnel of a reservoir will expand or contract due to temperature changes. To ensure the safety and functionality of the pipeline during operation, its actual pipe diameter value needs to be dynamically corrected through a formula; According to the temperature field distribution, the thermal expansion amount of each position of the pipeline is calculated section by section; For example, the inner wall area expands significantly due to high temperature, while the outer wall area expands less due to low temperature; Materials will expand or contract when the temperature changes; When the temperature rises, the pipeline will expand radially outward, resulting in an increase in the pipe diameter; When the temperature drops, the opposite is true. The average value of the inner and outer wall temperatures is used as the correction basis to avoid local expansion distortion caused by too high or too low temperature on one side; Correct the pipe diameter value: The original pipe diameter value is obtained by caliper or laser scanning, and the correction value is dynamically adjusted in combination with real-time feedback, such as pipeline deformation, temperature expansion coefficient.

[0020] For example, if the original pipe diameter is 100 mm and the diameter expands by 0.5% due to temperature rise, the corrected pipe diameter value is 100.5 mm; Generate telescopic instructions for the umbrella structure: The core of the fuzzy PID controller is to dynamically adjust the three parameters of the traditional PID, namely the proportional gain Kp, the integral gain Ki, and the derivative gain Kd, through fuzzy logic. Its input is the error and the rate of change of the error, and the output is the correction value of the PID parameters. Mamdani fuzzy inference is adopted to calculate the weight of each rule, and through the centroid method, the fuzzy output is converted into precise values of Kp, Ki, and Kd. According to the corrected values of Kp, Ki, and Kd, the telescopic amount is adjusted in real time to ensure that the nozzle / sensor always adheres to the pipe wall. The umbrella structure is driven by a servo motor or a hydraulic cylinder, generates specific instructions after receiving the telescopic amount signal, and at the same time sets the maximum telescopic speed and stroke range to avoid mechanical overload. Mamdani fuzzy inference: Convert the error and the rate of change of the error into the linguistic variable values of fuzzy sets, and determine the membership degrees of each linguistic variable according to the predefined membership functions. Based on the fuzzy rule base, calculate the weight of each rule through the Mamdani inference method, and aggregate all the rule outputs to obtain a fuzzy output set. Use the centroid method to convert the aggregated fuzzy output into precise numerical values of the proportional gain Kp, the integral gain Ki, and the derivative gain Kd for real-time adjustment of the telescopic instructions of the umbrella structure. Proportional gain Kp: Determines the proportional relationship between the controller output and the current error, that is, the difference between the target value and the actual value. Integral gain Ki: Affects the controller output by accumulating all past errors, mainly used to eliminate static errors, that is, the small errors that still exist when the system tends to be stable. Derivative gain Kd: Based on the rate of change of the error, predicts the future error trend, so as to take measures in advance to slow down the rate of error change.

[0021] The multi-process collaborative control module dynamically sets the spraying reference speed and relationship according to the pipe diameter curve and the coating thickness requirement. It uses the YOLOv5 model to detect foreign objects on the pipe wall in real time. When the height of the identified obstacle is greater than the height threshold c, it pauses the progress, triggers the cleaning, and then continues the process. If the obstacle still cannot be cleared, it marks the position and notifies the console for manual intervention. When the height of the identified obstacle is less than or equal to the height threshold c, it pauses the progress and triggers local supplementary grinding. Dynamically set the spraying reference speed: Through the determination of the spraying reference speed, based on the principle of conservation of paint flow, precise control is achieved for the dynamic balance of the pipe diameter, coating thickness, and paint flow. The real-time pipe diameter is obtained through laser scanning, and combined with the data of the temperature sensor, the thermal expansion amount of the pipe is calculated to obtain the corrected actual pipe diameter. The effective flow is adjusted according to the ambient temperature to offset the change in paint fluidity caused by high temperature. Calculate the volume of the coating required to reach the specified thickness per unit area based on the characteristics of the selected coating, such as the percentage of solid content, viscosity, etc.; set a reasonable spraying speed according to the output efficiency of the spraying equipment under different conditions and the length of the pipeline; ensure that the coating can evenly cover the surface of the pipeline; Real-time detection of foreign objects on the pipe wall: In YOLOv5, replace the original C3 module with the C2f module to enhance the ability to extract gradient flow information; the C2f module reduces the computational amount while retaining more detailed features through the residual structure and 1×1 convolution for dimensionality reduction, replaces the traditional CIoU loss function, solves the problem of inaccurate positioning of small sticky objects, and improves the regression accuracy of the bounding box; introduce channel attention or effective channel attention in the Backbone or Neck to enhance the attention to the key features of foreign objects and suppress background interference; For problems such as uneven illumination and noise interference that may exist inside the pipeline, the following technologies need to be combined for optimization; enhance the image contrast through contrast-limited adaptive histogram equalization to reduce the impact of environmental noise on detection; introduce BiFPN or FPN in the Neck part to achieve cross-scale feature fusion and adapt to the detection of foreign objects of different sizes; use YOLOv5 to reduce the number of parameters and computational amount and adapt to real-time operation on edge devices; For example: The vision sensor carried by the robot inside the pipeline collects the wall image in real time; perform CLAHE enhancement on the image to improve the contrast of low-light areas, remove noise and background interference; use Mosaic data augmentation and adaptive anchor box calculation to improve the generalization ability of the model to foreign objects, adapt the input image size to ensure that small objects can be effectively detected; output the detected foreign object category, the bounding box coordinates of the foreign object, as well as the confidence level and category probability; If the inner diameter of the pipeline wall and the camera parameters are known, the height of the obstacle can be calculated through the triangular relationship; according to the impact of the obstacle on spraying in past experiments, preset the obstacle height threshold c to trigger the warning system; at the same time, the position and height of the foreign object can be marked on the monitoring screen to assist the operation and maintenance personnel in making quick decisions; The inner wall of a water conveyance tunnel in a certain reservoir is scaled due to long-term erosion by water flow. It is necessary to monitor the scaling thickness in real time to prevent blockage; use the camera installed on the robot to collect images at intervals of every 5 meters; use the YOLOv5 model, train on the scaling sample data set, optimize the NWD loss function to improve the detection accuracy of small objects, introduce the CBAM attention mechanism to suppress the interference of water stain reflection, calculate the scaling height through geometric formulas, and when the scaling height exceeds 5mm, the system automatically triggers a prompt for manual intervention and records the position coordinates for subsequent processing; Trigger cleaning: Grind the foreign object with a grinding and rust-removing vehicle, and clean the residue after grinding with a dust-absorbing cleaning vehicle.

[0022] The coating thickness closed-loop control module obtains the thickness feedback in real time, compares it with the target value, adjusts the rotation speed of the spraying motor, and performs floating compensation on the nozzle; During the anti-corrosion process of industrial pipelines, the uniformity and accuracy of the coating thickness directly affect the subsequent service quality of the pipelines; traditional open spraying systems rely on preset parameters and are difficult to cope with environmental fluctuations, resulting in coating thickness deviations. Through real-time feedback, dynamic adjustment, and floating compensation, high-precision and high-consistency coating control is achieved; the electromagnetic eddy current is used to automatically detect the coating thickness, and the operator sets the target coating thickness through the HMI; if the measured value is lower than the target value, the spraying amount needs to be increased; if it is higher than the target value, the spraying amount needs to be reduced; the motor speed is dynamically adjusted according to the deviation size to eliminate long-term deviations, prevent the coating thickness from continuously deviating from the target, predict the deviation trend, and suppress the oscillation caused by the fluctuation of the workpiece moving speed; Perform floating compensation on the nozzle: By adjusting the nozzle attitude, ensure that the paint is evenly covered; integrate force sensors and displacement sensors to detect the contact force and relative position between the nozzle and the workpiece in real time. For example, when there is a protrusion on the inner surface of the pipeline, the sensor detects an increase in the contact force and triggers the nozzle to shift upward; automatically adjust the floating resistance of the nozzle according to the contact force to avoid excessive oscillation or hysteresis; detect uneven spraying on the pipe wall through YOLOv5, and the closed-loop system automatically increases the local coating thickness and adjusts the nozzle angle through floating compensation to cover irregular areas.

[0023] Embodiment 2 Please refer to Figure 2 , based on Embodiment 1, Embodiment 2 of the present application also provides a method for controlling a pipeline grinding robot, including the following specific steps: Step 1: Obtain the pipe diameter curve through continuous laser scanning, and splice the continuously scanned point cloud data in real time through the SLAM algorithm to construct a dynamic three-dimensional topological model of the pipeline; obtain the attitude data of the robot, calculate the offset of the robot on the central axis inside the pipeline; obtain the monitored environmental temperature inside the pipeline, construct a dynamic thermal expansion model of the pipeline through finite element analysis of heat conduction, and correct the pipe diameter value according to the thermal expansion model; according to the offset and the corrected pipe diameter value, calculate the telescopic amount of the umbrella structure through a fuzzy PID controller, and generate an instruction for the telescopic amount of the umbrella structure; Step 2: Dynamically set the spraying reference speed and relationship according to the pipe diameter curve and the coating thickness requirement; detect foreign objects on the pipe wall in real time through the YOLOv5 model. When the height of the identified obstacle is greater than the height threshold c, pause the progress, trigger the clearance, and continue the process; if the clearance still cannot be achieved, mark the position and notify the console for manual intervention; When the height of the identified obstacle is less than or equal to the height threshold c, pause the progress and trigger local supplementary grinding; Step 3: Obtain the thickness feedback in real time, compare it with the target value, adjust the rotation speed of the spraying motor, and perform floating compensation on the nozzle.

[0024] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those of ordinary skill in the art will recognize that the units and algorithm steps of the examples described in connection with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution.

[0025] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the objectives of the solution of this embodiment.

[0026] As described above, only the specific implementation manners of the present application are provided, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application.

Claims

1. A system for controlling a pipeline grinding robot, characterized in that, The system includes: A variable-diameter adaptive preprocessing module that obtains the pipe diameter curve through continuous laser scanning, stitches the continuously scanned point cloud data in real time through the SLAM algorithm to construct a dynamic three-dimensional topological model of the pipeline; obtains the attitude data of the robot, calculates the offset of the robot's central axis in the pipeline; obtains the monitored environmental temperature in the pipeline, constructs a dynamic thermal expansion model of the pipeline through finite element analysis of heat conduction, and corrects the pipe diameter value according to the thermal expansion model; according to the offset and the corrected pipe diameter value, calculates the telescopic amount of the umbrella structure through a fuzzy PID controller to generate an umbrella structure telescopic instruction; A multi-process collaborative control module that dynamically sets the spraying reference speed and relationship according to the pipe diameter curve and the coating thickness requirement; detects foreign objects on the pipe wall in real time through the YOLOv5 model. When the height of the identified obstacle is greater than the height threshold c, it pauses the progress, triggers cleaning, and continues the process; if it still cannot be cleared, marks the position and notifies the console for manual intervention; When the height of the identified obstacle is less than or equal to the height threshold c, it pauses the progress and triggers local supplementary grinding; A coating thickness closed-loop control module that obtains the thickness feedback in real time, compares it with the target value, adjusts the rotation speed of the spraying motor, and performs floating compensation on the nozzle.

2. The system for controlling a pipeline grinding robot according to claim 1, wherein The process of obtaining the pipe diameter curve is as follows: Perform circular arc fitting on the stitched point cloud data, calculate the inner diameter value of each cross-section; identify the variable-diameter area through curvature calculation to generate a segmented continuous pipe diameter curve.

3. The system for controlling a pipeline grinding robot according to claim 2, wherein, The process of real-time stitching through the SLAM algorithm is as follows: Apply the LIO-SLAM algorithm to achieve dynamic stitching of point cloud data through the following steps: S101: Use IMU data, accelerometers and gyroscopes to calculate the pose change of the robot in real time; S102: Align the point cloud data of adjacent frames through the ICP algorithm; S103: Based on the SLAM framework, construct the stitched point cloud data into a dynamic three-dimensional topological model of the pipeline; Use voxel filtering to reduce the point cloud density while retaining geometric features; remove noise points through statistical outlier detection.

4. The system for controlling a pipeline grinding robot according to claim 3, characterized in that, The process of constructing the dynamic three-dimensional topological model of the pipeline is as follows: Generate the inner wall of the pipeline based on the point cloud data; each frame of point cloud data contains spatial point coordinate information, and the point cloud is filtered to remove isolated points; Use the ICP algorithm to match a newly acquired frame of point cloud with the previously constructed map, find the rigid body transformation between the two frames of point cloud, and form a global map by integrating the results of multiple local registrations; based on the registered point cloud data, use computer vision algorithms to extract the key geometric features inside the pipeline and update them dynamically.

5. A system for controlling a pipeline grinding robot according to claim 1, characterized in that, The process of calculating the offset of the robot's central axis in the pipeline is as follows: Obtain the angular velocity and acceleration of the robot body through sensors, extract the central axis of the pipeline based on the three-dimensional topological model of the pipeline constructed by laser scanning as the ideal trajectory; combine the attitude angle data of the IMU and the robot kinematic model to calculate the actual central axis position of the robot; convert the attitude angle measured by the IMU into the pose in the robot body coordinate system, and obtain the translational displacement of the robot through the odometer; compare the actual central axis position with the ideal trajectory to calculate the deviation.

6. The system for controlling a pipeline grinding robot according to claim 5, wherein, The process of correcting the pipe diameter value is as follows: Based on the properties of the pipeline material, the model is meshed, the continuous pipeline structure is discretized into many small cells, and appropriate boundary conditions are set according to the actual situation. The finite element method is used to solve the heat conduction equation to predict the temperature field distribution of each part of the pipeline over time under given conditions; According to the temperature field distribution, the thermal expansion of each position of the pipeline is calculated section by section; the average value of the inner and outer wall temperatures is used as the correction basis.

7. A system for controlling a pipeline grinding robot according to claim 6, characterized in that, The specific process of generating the telescopic instruction for the umbrella structure is as follows: The three parameters of the traditional PID, the proportional gain Kp, the integral gain Ki, and the derivative gain Kd, are dynamically adjusted through fuzzy logic. Its input is the error and the error change rate, and the output is the PID parameter correction value; Mamdani fuzzy inference is used to calculate the weight of each rule, and through the centroid method, the fuzzy output is converted into accurate Kp, Ki, and Kd values. According to the corrected Kp, Ki, and Kd values, the telescopic amount is adjusted in real time.

8. A system for controlling a pipeline grinding robot according to claim 1, characterized in that, The process of dynamically setting the spraying reference speed and relationship is as follows: Through the determination of the spraying reference speed, based on the principle of conservation of paint flow, precise control is achieved for the dynamic balance of the pipe diameter, coating thickness, and paint flow; the real-time pipe diameter is obtained through laser scanning, combined with the data of the temperature sensor, the thermal expansion of the pipeline is calculated, and the corrected actual pipe diameter is obtained; the effective flow is adjusted according to the ambient temperature to offset the change in paint fluidity caused by high temperature; The spraying speed is set according to the output efficiency of the spraying equipment under different conditions.

9. A system for controlling a pipeline grinding robot according to claim 1, characterized in that, The process of the nozzle for floating compensation is as follows: The electromagnetic eddy current is used to automatically detect the coating thickness, and the operator sets the target coating thickness through the HMI; the motor speed is dynamically adjusted according to the deviation size to eliminate the long-term deviation. The force sensor and displacement sensor are integrated to detect the contact force and relative position between the nozzle and the workpiece in real time; the floating resistance of the nozzle is automatically adjusted according to the contact force, and the nozzle angle is adjusted through floating compensation.

10. A method for controlling a pipeline grinding robot, characterized in that It includes the following steps: Step 1: Continuously scan the pipe diameter curve through the laser, and the point cloud data obtained from the continuous scan is stitched in real time through the SLAM algorithm to construct a dynamic three-dimensional topological model of the pipeline; obtain the pose data of the robot, and calculate the offset of the robot's central axis in the pipeline; obtain the monitored ambient temperature in the pipeline, construct a dynamic thermal expansion model of the pipeline through finite element analysis of heat conduction, and correct the pipe diameter value according to the thermal expansion model; according to the offset and the corrected pipe diameter value, calculate the telescopic amount of the umbrella structure through the fuzzy PID controller, and generate the telescopic instruction for the umbrella structure; Step 2: Dynamically set the spraying reference speed and relationship according to the pipe diameter curve and the coating thickness requirement; use the YOLOv5 model to detect foreign objects on the pipe wall in real time. When the height of the identified obstacle is greater than the height threshold c, the progress is paused, the cleaning is triggered, and the process continues; if the cleaning still cannot be completed, mark the position and notify the console for manual intervention; When the height of the identified obstacle is less than or equal to the height threshold c, the progress is paused, and local supplementary grinding is triggered; Step 3: Obtain the thickness feedback in real time, compare it with the target value, adjust the spraying motor speed, and perform floating compensation on the nozzle.

Images