Coal mine drainage pipeline mounting robot and working method thereof

Abstract

The invention discloses a coal mine drainage pipeline mounting robot and a working method thereof, and belongs to the field of underground pipeline mounting, the coal mine drainage pipeline mounting robot comprises a robot body, and a magnetic attraction type crawler climbing mechanism, a pipeline tensioning mechanism, a sealing mechanism, a multi-source environment sensing mechanism and an intelligent control mechanism which are integrated on the robot body, the multi-source environment sensing mechanism is electrically connected with the intelligent control mechanism, and the intelligent control mechanism is electrically connected with the magnetic attraction type crawler climbing mechanism and the pipeline tensioning mechanism. By the adoption of the coal mine drainage pipeline installation robot and the working method thereof, automatic, high-precision and safe installation of underground pipelines is achieved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of downhole pipeline installation, in particular to a coal mine drainage pipeline installation robot and a working method thereof. BACKGROUND

[0002] The coal mine drainage pipeline is a key infrastructure for the safety production of coal mines, and the installation quality directly affects the downhole drainage efficiency and operation safety. The traditional coal mine drainage pipeline installation relies on manual operation, and the following core problems exist: 1. The space in the coal mine is narrow and the environment is harsh (high humidity, dust, high anti-explosion requirement), manual installation has high labor intensity and poor safety, and safety accidents such as falling and collision are prone to occur; 2. The positioning accuracy is low when the pipes are connected manually, and the coaxiality deviation is prone to occur, which leads to increased water leakage and local water head loss, and affects the drainage efficiency; 3. The vibration interference is obvious in the downhole, and the traditional installation equipment lacks an effective vibration suppression mechanism, which further reduces the installation accuracy; 4. The existing pipeline installation robots are mostly suitable for conventional environments, and have defects such as poor path planning adaptability, insufficient climbing stability and inconvenient human-computer interaction, and are difficult to adapt to the complex curved pipeline in the coal mine and the anti-explosion requirement. SUMMARY

[0003] The application aims to provide a coal mine drainage pipeline installation robot and a working method thereof, which solve the above technical problems.

[0004] To achieve the above-mentioned purpose, the application provides a coal mine drainage pipeline installation robot, which comprises a robot body, a magnetic track climbing mechanism, a pipeline tensioning mechanism, a sealing mechanism, a multi-source environment sensing mechanism and an intelligent control mechanism integrated on the robot body, the multi-source environment sensing mechanism is electrically connected with the intelligent control mechanism, the intelligent control mechanism is electrically connected with the magnetic track climbing mechanism and the pipeline tensioning mechanism respectively, and is used for realizing the track differential steering and electromagnetic adsorption point distributed on-off collaborative climbing of the complex curved pipeline with a diameter of 200 mm in the coal mine 50- 200mm, combining the three-dimensional path planning and ToF-vision fusion accurate positioning of the MOA* algorithm fusion track climbing characteristics carried on the intelligent control mechanism, inhibiting the residual vibration of the track climbing and the downhole environmental interference through the ADRC-fractional order fusion control, and driving the pipeline tensioning mechanism to complete the centering tensioning, attitude compensation and micron-level connection.

[0005] A working method of a coal mine drainage pipeline installation robot, comprising the following steps: S1, transporting the pipeline to be installed to the end of the installed pipeline in the coal mine installation area, and locating the pipeline to be installed in the operating range of the robot; Meanwhile, the robot is controlled by gestures to enter the installed pipeline from the other end of the installed pipeline, a multi-source environment perception mechanism is started, three-dimensional distance data of the inner wall of the installed pipeline, pipeline images and depth information are synchronously collected by a ToF sensor and an RGB-D camera, and a vibration sensor is used to collect a vibration reference value of the downhole environment, and after preprocessing, a comprehensive data set of fused environment point cloud data and pipeline semantic segmentation results is obtained; S2, based on the comprehensive data set output by S1 and pipeline design parameters, a three-dimensional grid space is constructed and a path search boundary is defined, a MOA* path planning algorithm with fused crawler climbing characteristics is used, a dynamic cost function including an axis deviation weight, an obstacle avoidance weight and a climbing cost weight is designed, invalid path branches are removed through a dynamic probability pruning strategy, and an optimal path and a corresponding crawler walking parameter sequence are generated; S3, based on the optimal path and the crawler walking parameters output by S2, distributed electromagnetic adsorption points on the outer surface of the crawler are activated, the total adsorption force meets the dual constraints of anti-sliding and anti-overturning, and climbing is realized by using a crawler driving unit, and in the climbing process, a step switching strategy for adsorption points in the front half and the rear half is adopted to ensure adsorption continuity; S4, based on the climbing position data output by S3, when the robot reaches the installation position, the inner wall of the pipeline to be installed is tensioned by using a three-jaw tensioning disc, and an ADRC-fractional order fusion vibration suppression algorithm is started, residual vibration of the crawler climbing and downhole environmental interference are estimated in real time by using an extended state observer, and a vibration suppression control amount is generated; S5, based on the vibration suppression effect of S4, axial displacement adjustment is performed according to a closed-loop strategy, deviations in roll and pitch directions are compensated by using an attitude adjustment platform, a pipe butt joint gap sensor is used to monitor the pipe butt joint gap in real time, and a RGB-D camera is used to dynamically calculate coaxiality, and repeated adjustment is performed until the butt joint gap is not greater than 0.2mm and the coaxiality is not greater than 0.1mm, and pipe butt joint is completed; S6, based on the butt joint result of S5, defect features of the pipe butt joint area are extracted by semantic segmentation, a gap width and a displacement are calculated, a pipe-pipe interaction SPH numerical model is established, pipe material parameters and defect features are input to calculate a maximum stress, and a quality risk level is determined according to a relationship between the maximum stress and a material allowable stress, if the quality risk level is qualified, a butt joint completion signal is output, and if the quality risk level is unqualified, a closed-loop rework instruction is triggered; S7, when the qualified result is output by S6, a sealant gun is used to inject sealant between the pipeline to be installed and the installed pipeline, the sealant is used to fill the gap between the pipelines by gravity flow, and the pipelines are sealed by solidification; S8, the robot releases the pipeline clamp and prepares for the next installation position climbing according to an optimal path update strategy, and after completing all pipeline installation tasks, the robot returns to the starting position and waits.

[0006] Therefore, the beneficial effects of the above-mentioned coal mine drainage pipeline installation robot and its working method are as follows: 1. Significantly improves operational safety: No need for manual operation in the narrow, humid, and dusty underground environment; robot control is achieved through remote gesture interaction, avoiding safety accidents such as falls and collisions, reducing labor intensity, and meeting underground explosion-proof requirements; 2. Significantly improve installation accuracy: Through ToF-vision fusion positioning, attitude compensation adjustment and closed-loop control strategies, micron-level docking with a gap of ≤0.2mm and coaxiality of ≤0.1mm is achieved, solving problems such as large coaxiality deviation and water leakage in manual installation and improving drainage efficiency; 3. Effectively suppress environmental interference: The ADRC-fractional order fusion control algorithm is adopted to estimate and suppress residual vibration of track climbing and interference from the underground environment in real time, overcoming the defect of poor vibration adaptability of traditional equipment and ensuring installation stability in complex environments; 4. Enhance path planning adaptability: Based on the MOA* algorithm and the track climbing characteristics, a dynamic cost function and probabilistic pruning strategy are designed to generate the optimal path, which can be adapted to the complex curved surface and obstacle environment of underground pipelines and improve climbing stability (through the step-by-step switching of distributed electromagnetic adsorption points and the dual constraints of anti-slip and anti-overturning). 5. Achieve closed-loop quality control: Extract defect features through semantic segmentation, calculate the maximum stress and determine the quality risk level by combining the SPH numerical model, trigger rework instructions for non-conformities, and ensure sealing effect through sealant injection and curing to ensure reliable installation quality. 6. Optimize human-computer interaction experience: Equipped with a remote gesture interaction module, it can recognize the operator's dynamic gestures and convert them into control commands, eliminating the need for complex wired connections, improving the ease of operation, and adapting to the operational needs of complex underground working scenarios.

[0007] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the structure of a coal mine drainage pipeline installation robot according to the present invention.

[0009] Figure Labels 1. Installed pipe; 2. Robot body; 3. Sealing gun; 4. Pipe to be installed; 5. Three-jaw tensioning disc; 6. Axial displacement adjustment unit; 7. Attitude adjustment platform; 8. Electromagnetic adsorption point; 9. Track drive unit; 10. RGN-D camera. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the embodiments of the present invention and are not intended to limit the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of this application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout.

[0011] It should be noted that the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, such as a process, method, system, product, or server that includes a series of steps or units, not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such process, method, product, or device.

[0012] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0013] like Figure 1 As shown, a coal mine drainage pipeline installation robot includes a robot body 2 and a magnetic tracked climbing mechanism, a pipeline tensioning mechanism, a sealing mechanism, a multi-source environmental sensing mechanism, and an intelligent control mechanism integrated on the robot body 2. The multi-source environmental sensing mechanism is electrically connected to the intelligent control mechanism, and the intelligent control mechanism is electrically connected to both the magnetic tracked climbing mechanism and the pipeline tensioning mechanism, respectively, for use in underground coal mine drainage pipeline installation. 50- The 200mm complex curved pipe features differential steering of the tracked vehicle and 8 distributed on / off points for coordinated climbing. It also incorporates the MOA* algorithm on the intelligent control mechanism to integrate the tracked vehicle climbing characteristics into 3D path planning and ToF-vision fusion for precise positioning. Through ADRC-fractional order fusion control, it suppresses residual vibration of the tracked vehicle climbing and interference from the underground environment, driving the pipe tensioning mechanism to complete centering tensioning, attitude compensation, and micron-level docking.

[0014] Specifically, the magnetic track climbing mechanism includes a track drive unit 9 set on the robot body 2 and distributed electromagnetic adsorption points 8 embedded in the track body of the track drive unit 9 in a matrix. Each electromagnetic adsorption point 8 includes a permanent magnet embedded in the track body and an electromagnetic coil wound around the permanent magnet, which are used to generate adsorption force to adsorb the inner wall of the pipe. The pipeline tensioning mechanism includes an attitude adjustment platform 7 fixed to the front end of the robot body 2, an axial displacement adjustment unit 6 fixed to the output end of the attitude adjustment platform 7, and a three-jaw tensioning disc set at the output end of the axial displacement adjustment unit 6. The axial displacement adjustment unit is an electric telescopic structure. The sealing mechanism includes a sealing glue gun 3 fixed to the top of the front end of the robot body 2; The multi-source environmental perception mechanism includes a ToF sensor and an RGN-D camera 10 installed at the front end of the robot body 2, a vibration sensor, a gap sensor, and a force sensor installed on the three-jaw tensioning plate, a piezoelectric sensor installed at the electromagnetic adsorption point 8, and an attitude sensor. The ToF sensor and the RGN-D camera 10 are used to acquire three-dimensional distance data and image data of the inner wall of the pipe, respectively. The vibration sensor and the gap sensor are used to acquire vibration data and pipe installation gap, respectively. The force sensor is used to collect the tension force of the three-jaw tensioning plate. The piezoelectric sensor is used to collect the adsorption force of the corresponding electromagnetic adsorption point 8. The attitude sensor is used to collect the robot's roll angle and pitch angle.

[0015] The invention also includes a remote gesture interaction module, which is used to recognize the operator's dynamic gestures and convert them into control commands that are transmitted to the intelligent control mechanism to achieve remote control.

[0016] A method for operating a robot for installing drainage pipes in coal mines includes the following steps: S1. Transfer the pipe 54 to be installed to one end of the underground installation area near the already installed pipe 1, and within the robot's operating range; Simultaneously, the robot is controlled by gestures to enter the interior of the installed pipe 1 from the other end of the installed pipe 1, and the multi-source environmental perception mechanism is activated. The three-dimensional distance data of the inner wall of the installed pipe 1 and the pipe image and depth information are collected synchronously through the ToF sensor and RGB-D camera. The vibration sensor is used to collect the underground environmental vibration benchmark value. After preprocessing, a comprehensive dataset is obtained by fusing environmental point cloud data and pipe semantic segmentation results. S2. Based on the comprehensive dataset and pipeline design parameters output by S1, a three-dimensional mesh space is constructed and the path search boundary is defined. The MOA* path planning algorithm, which integrates track climbing characteristics, is adopted. A dynamic cost function is designed, which includes axis deviation weight, obstacle avoidance weight and climbing cost weight. Invalid path branches are removed through dynamic probabilistic pruning strategy to generate the optimal path and the corresponding track walking parameter sequence. S3. Based on the optimal path and track walking parameters output by S2, activate the distributed electromagnetic adsorption points 8 on the outer surface of the track to make the total adsorption force meet the dual constraints of anti-slipping and anti-overturning, and use the track drive unit 9 to achieve climbing. During the climbing process, a step-by-step switching strategy of the adsorption points in the first half and the second half is adopted to ensure the continuity of adsorption. S4. Based on the climbing position data output by S3, determine when the robot reaches the installation position, use the three-jaw tensioning disc to tension the inner wall of the pipe 54 to be installed, and start the ADRC-fractional order fusion vibration suppression algorithm. Through the extended state observer, estimate the residual vibration of the track climbing and the interference of the underground environment in real time, and generate the vibration suppression control quantity. S5. Based on the vibration suppression effect of S4, axial displacement is adjusted according to the closed-loop strategy, and the attitude adjustment platform 7 is used to compensate for roll and pitch deviations. At the same time, the docking gap sensor monitors the pipe docking gap in real time, and the RGB-D camera dynamically calculates the coaxiality. The adjustment is repeated until the docking gap is no greater than 0.2mm and the coaxiality is no greater than 0.1mm, thus completing the pipe docking. S6. Based on the docking results of S5, extract the defect features of the pipe docking area through semantic segmentation, calculate the gap width and offset, establish a pipe-to-pipe interaction SPH numerical model, input the pipe material parameters and defect features to calculate the maximum stress, and determine the quality risk level according to the relationship between the maximum stress and the allowable stress of the material. If it is qualified, output the docking completion signal; if it is unqualified, trigger the closed-loop rework command. S7. When S6 outputs a qualified result, use the sealant gun 3 to inject sealant between the pipe to be installed 54 and the installed pipe 1. The sealant will fill the gap between the pipes by gravity and will be cured and sealed between the pipes. S8. The robot releases the pipe clamp and updates its strategy according to the optimal path to prepare for climbing to the next installation position. After completing all pipe installation tasks, the robot returns to the starting position to standby.

[0017] Step S1 specifically includes the following steps: S11, Utilizing a ToF sensor Simultaneously acquire 3D distance data of the inner wall of the installed pipe 1 and pipe images with an RGB-D camera. With depth information Then, the pixel coordinates are converted to world coordinates through coordinate transformation to obtain the blended environment point cloud. and the initial positioning frame of the pipeline ,in, Indicates the fusion of environmental point clouds Three-dimensional world coordinates; Indicates the initial positioning box The minimum coordinate value; Indicates the initial positioning box The maximum coordinate value; S12. Extract the pipeline region using the Deeplabv3+ semantic segmentation algorithm: ; In the formula, Indicates the total loss; Indicates the weighting coefficient; Represents cross-entropy loss; Represents the distance loss function; S13. Perform ellipse fitting on the segmented pipe point cloud to obtain the central axis of the pipe: ; In the formula, A parametric curve representing the central axis of the pipeline; and These represent the major and minor semi-axises after fitting the ellipse to the pipe cross-section; The parameter variables representing the parametric curve of the center axis; This represents the slope of the central axis.

[0018] The expression for the dynamic cost function mentioned in step S2 is as follows: ; In the formula, , and These represent the axis deviation weight, obstacle avoidance weight, and climbing cost weight, respectively. Representing path points To the center axis of the pipeline The distance; Representing path points The cost of obstacles at the location; Representing path points Compared with the previous path point Path curvature between; The MOA* algorithm, with its dynamic cost function at its core, searches all path points to obtain the optimal path. Total path cost and the corresponding track movement parameter sequence, wherein the track movement parameters include the target climbing speed for each segment of the path. and steering angle .

[0019] Step S3 specifically includes the following steps: S31. Calculate the total adsorption force of the target. : ; in, ; ; ; In the formula, Indicates the safety factor of the adsorption force; This represents the minimum threshold of total adsorption force; This represents the minimum total adsorption force against sliding constraints; This represents the minimum total adsorption force required to resist overturning. Indicates the coefficient of friction between the track and the installed inner wall; Indicates the inner radius of the installed pipe 1; This represents the robot's total weight. The horizontal lever arm representing the distance from the point of application of the adsorption force to the robot's center of gravity; S32. Based on the target total adsorption force, calculate the adsorption force required for a single set of electromagnetic adsorption points 8. and corresponding drive current and driving voltage : ; ; ; In the formula, This indicates the number of electromagnetic adsorption points 8 that are in an active state; This indicates the distance between electromagnetic adsorption point 8 and the air gap inside the pipe wall; Indicates the permeability of free space; This represents the effective area of ​​the adsorption surface of a single set of adsorption points; This indicates the number of turns of the electromagnetic coil at a single adsorption point; This represents the resistance of a single set of adsorption point coils; S33, Based on Optimal Path and track walking parameters, configure climbing speed Target speed of track motor And the rotational speeds on the inner and outer sides of the pipe bends: ; ; ; ; In the formula, Indicates the diameter of the track drive wheel; Indicates the actual speed of the drive motor; Indicates the transmission ratio of the harmonic reducer; This indicates the number of steps per revolution of the track drive motor; and These represent the speeds of the outer and inner tracks during turning, respectively. Indicates the center-to-center distance between the two tracks; Indicates the steering adjustment time; During the climb, a step-by-step strategy is implemented: activation of the first half of the adsorption points, transition switching, and activation of the second half of the adsorption points. The trigger distance for adsorption point switching is specified. Switching interval time and the adsorption force retention threshold during the switching process The calculation formula is as follows: ; ; ; In the formula, Indicates track pitch; Furthermore, during the climb, at each waypoint... Pause for 0.5 seconds, and use the attitude sensor results to correct the current adjustment amount based on the adsorption force of a single adsorption point. and attitude correction speed difference adjustment amount : ; ; in, ; ; In the formula, Indicates the attitude deviation value; This indicates the deviation value of the adsorption force; Indicates the robot's roll angle deviation; Indicates the robot's pitch angle deviation; This represents the actual total adsorption force (which is obtained by summing the adsorption forces of individual adsorption points). During the climb, ToF sensors and RGB-D cameras were used to collect real-time 3D distance data of the pipe's inner wall, as well as pipe images and depth data, to determine the climbing position. Simultaneously calculate the positioning error : ; ; In the formula, and These represent the ToF and RGB-D weights, respectively. Indicates the relative position based on the ToF sensor's own coordinate system; This indicates the pixel-depth correspondence based on the RGB-D camera's own coordinate system; This represents the three-dimensional world coordinates corresponding to the robot's actual climbing position after calibration. This represents the three-dimensional world coordinates corresponding to the target installation station that the robot is climbing; The final output is the corrected real-time climbing position. and positioning error and in positioning error At this time, the three-dimensional deviation data between the current climbing position and the target installation position is re-acquired, and the track axial displacement and differential steering angle are adjusted based on the three-dimensional deviation data until the requirements are met. .

[0020] Step S4 specifically includes the following steps: S41, when the corrected real-time climbing position When the pipe reaches 50mm from the end face of the installed pipe 1, stop climbing and use the electromagnetic adsorption point 8 for positioning. At this time, the three-jaw tensioning disc extends out of the installed pipe 1 and drives the three-jaw tensioning disc to extend synchronously to tension the outer wall of the pipe to be installed 54 until the actual tension force collected by the force sensor is equal to the target tension force. At this time, the axis of the pipe to be installed 54 coincides with the axis of the installed pipe 1. S42. Activate the ADRC-fractional-order fusion vibration suppression algorithm to construct a coupled vibration dynamic model of the three-jaw tensioning disc, the pipe to be installed 54, and the tracked motor: ; In the formula, This indicates the total mass of the three-jaw tensioning disc, the pipe to be installed 54, and the track motor; and These represent the system's damping coefficient and stiffness coefficient, respectively. , and These represent the vibration displacement, vibration velocity, and vibration acceleration along the pipeline axis, respectively. Indicates the disturbance force of the downhole environment; This indicates residual vibration interference force during track climbing; S43. Estimate the real-time total disturbance using a nonlinear observer based on a coupled vibration dynamics model. and will real-time total interference Decomposed into residual vibration interference force during track climbing and the disturbance force of the downhole environment Then, the disturbance force of the underground environment After elimination, the final total interference is obtained. : ; S44, Generating vibration suppression control quantity : ; in, ; In the formula, Indicates the basic control quantity; This represents the control gain coefficient; , , These represent the proportional, differential, and integral coefficients, respectively. Indicates vibration displacement deviation; Indicates the order of integration; This represents a fractional differential operator.

[0021] In step S5, based on the vibration suppression control quantity Adjust the axial adjustment mechanism and the attitude adjustment mechanism to adjust the coaxiality between the pipe to be installed 54 and the installed pipe 1 until the final coaxiality is obtained using video detection technology. .

[0022] Step S6 specifically includes the following steps: S61. Perform semantic segmentation on the docking area, extract defect features, and calculate the gap width through curve fitting. : ; In the formula, and These represent the pixel coordinates on both sides of the gap to be detected; S62. Establish the SPH model for tube-to-tube interactions: ; In the formula, express; express; express; express; express; express; express; express; S63. For particles in the SPH model, their strain rate tensor The discrete form is: ; In the formula, Represents particles The quality; and They represent particles respectively and particles The speed (contact speed of the pipes after docking); and They represent particles respectively and particles The density; and They represent particles respectively and particles Position vector; Represents the sliding kernel function The gradient; S64. Obtain the particle by integrating the strain rate tensor. strain tensor : ; In the formula, Let represent the initial strain tensor, and ; S65. Calculation of stress tensor based on linear elastic constitutive equation : ; in, ; ; In the formula, and Both represent Lamé constants; Represents the trace of the strain tensor, and , , and They represent particles respectively Normal strain in the x, y, and z directions; Represents the Kronecker function; This indicates the elastic modulus of the pipe material; This indicates the Poisson's ratio of the pipe material; S66, stress tensor Converted to von Mises equivalent force : ; In the formula, , , These represent the normal stresses in the pipe along the x, y, and z directions, respectively. , , These represent the shear stresses of the pipe in the xy plane, yz plane, and zx plane, respectively. S67. Iterate through the equivalent stress of all pipe particles in the SPH model and take the maximum value as the maximum stress. : ; S68, Based on maximum stress Determine the risk level : ; In the formula, This indicates the allowable stress of the pipe material.

[0023] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A robot for installing drainage pipes in coal mines, characterized in that: The system includes a robot body and integrated on the robot body a magnetic track climbing mechanism, a pipe tensioning mechanism, a sealing mechanism, a multi-source environmental sensing mechanism, and an intelligent control mechanism. The multi-source environmental sensing mechanism is electrically connected to the intelligent control mechanism, and the intelligent control mechanism is electrically connected to both the magnetic track climbing mechanism and the pipe tensioning mechanism. This system is used to realize underground coal mining operations. 50- The 200mm complex curved pipe features differential steering of the tracked vehicle and distributed on / off electromagnetic adsorption points for coordinated climbing. It also incorporates the MOA* algorithm on the intelligent control mechanism to integrate the tracked vehicle climbing characteristics into 3D path planning and ToF-vision fusion for precise positioning. Through ADRC-fractional order fusion control, it suppresses residual vibration of the tracked vehicle climbing and interference from the underground environment, driving the pipe tensioning mechanism to complete centering tensioning, attitude compensation, and micron-level docking.

2. The coal mine drainage pipeline installation robot according to claim 1, characterized in that: The magnetic track climbing mechanism includes a track drive unit set on the robot body and distributed electromagnetic adsorption points embedded in the track body of the track drive unit in a matrix. Each electromagnetic adsorption point includes a permanent magnet embedded in the outer periphery of the track body and an electromagnetic coil wound around the permanent magnet, which are used to generate adsorption force to adsorb the inner wall of the pipe. The pipeline tensioning mechanism includes an attitude adjustment platform fixed to the front end of the robot body, an axial displacement adjustment unit fixed to the output end of the attitude adjustment platform, and a three-jaw tensioning disc set at the output end of the axial displacement adjustment unit. The axial displacement adjustment unit is an electrically telescopic structure. The sealing mechanism includes a sealing glue gun fixed to the top of the front end of the robot body; The multi-source environmental perception mechanism includes a ToF sensor and an RGN-D camera mounted on the front of the robot body, a vibration sensor, a gap sensor, and a force sensor mounted on the three-jaw tensioning plate, a piezoelectric sensor mounted on the electromagnetic adsorption point, and an attitude sensor. The ToF sensor and RGN-D camera are used to acquire three-dimensional distance data and image data of the inner wall of the pipe, respectively. The vibration sensor and gap sensor are used to acquire vibration data and pipe installation gap, respectively. The force sensor is used to collect the tension force of the three-jaw tensioning plate. The piezoelectric sensor is used to collect the adsorption force of the corresponding electromagnetic adsorption point. The attitude sensor is used to collect the robot's roll angle and pitch angle.

3. The coal mine drainage pipeline installation robot according to claim 2, characterized in that: It also includes a remote gesture interaction module, which is used to recognize the operator's dynamic gestures and convert them into control commands that are transmitted to the intelligent control mechanism to achieve remote control.

4. The working method of the coal mine drainage pipeline installation robot as described in claim 3, characterized in that: Includes the following steps: S1. Transport the pipeline to be installed to the underground installation area near one end of the already installed pipeline, within the robot's operating range; Simultaneously, the robot is controlled by gestures to enter the interior of the installed pipe from the other end of the pipe, and the multi-source environmental perception mechanism is activated. The ToF sensor and RGB-D camera are used to simultaneously collect three-dimensional distance data of the inner wall of the installed pipe and pipe images and depth information. The vibration sensor is used to collect the underground environmental vibration benchmark value. After preprocessing, a comprehensive dataset is obtained by fusing environmental point cloud data and pipe semantic segmentation results. S2. Based on the comprehensive dataset and pipeline design parameters output by S1, a three-dimensional mesh space is constructed and the path search boundary is defined. The MOA* path planning algorithm, which integrates track climbing characteristics, is adopted. A dynamic cost function is designed, which includes axis deviation weight, obstacle avoidance weight and climbing cost weight. Invalid path branches are removed through dynamic probabilistic pruning strategy to generate the optimal path and the corresponding track walking parameter sequence. S3. Based on the optimal path and track walking parameters output by S2, activate the distributed electromagnetic adsorption points on the outer surface of the track to ensure that the total adsorption force meets the dual constraints of anti-slipping and anti-overturning. Climbing is achieved using the track drive unit. During the climbing process, a step-by-step switching strategy between the adsorption points in the first half and the second half is adopted to ensure the continuity of adsorption. S4. Based on the climbing position data output by S3, determine when the robot reaches the installation position, use the three-jaw tensioning disc to tension the inner wall of the pipe to be installed, and start the ADRC-fractional order fusion vibration suppression algorithm. Through the extended state observer, estimate the residual vibration of the track climbing and the interference of the underground environment in real time, and generate the vibration suppression control quantity. S5. Based on the vibration suppression effect of S4, axial displacement is adjusted according to a closed-loop strategy, and the attitude adjustment platform is used to compensate for roll and pitch deviations. At the same time, the docking gap sensor monitors the pipe docking gap in real time, and the RGB-D camera dynamically calculates the coaxiality. The adjustment is repeated until the docking gap is no greater than 0.2mm and the coaxiality is no greater than 0.1mm, thus completing the pipe docking. S6. Based on the docking results of S5, extract the defect features of the pipe docking area through semantic segmentation, calculate the gap width and offset, establish a pipe-to-pipe interaction SPH numerical model, input the pipe material parameters and defect features to calculate the maximum stress, and determine the quality risk level according to the relationship between the maximum stress and the allowable stress of the material. If it is qualified, output the docking completion signal; if it is unqualified, trigger the closed-loop rework command. S7. When S6 outputs a qualified result, use a sealant gun to inject sealant between the pipe to be installed and the installed pipe. The sealant will flow by gravity to fill the gap between the pipes and then cure and seal the gap between the pipes. S8. The robot releases the pipe clamp and updates its strategy according to the optimal path to prepare for climbing to the next installation position. After completing all pipe installation tasks, the robot returns to the starting position to standby.

5. The working method of a coal mine drainage pipeline installation robot according to claim 4, characterized in that: Step S1 specifically includes the following steps: S11, Utilizing a ToF sensor Simultaneously acquire 3D distance data of the inner wall of the installed pipe and pipe images with an RGB-D camera. With depth information Then, the pixel coordinates are converted to world coordinates through coordinate transformation to obtain the blended environment point cloud. and the initial positioning frame of the pipeline ,in, Indicates the fusion of environmental point clouds Three-dimensional world coordinates; Indicates the initial positioning box The minimum coordinate value; Indicates the initial positioning box The maximum coordinate value; S12. Extract the pipeline region using the Deeplabv3+ semantic segmentation algorithm: ； In the formula, Indicates the total loss; Indicates the weighting coefficient; Represents cross-entropy loss; Represents the distance loss function; S13. Perform ellipse fitting on the segmented pipe point cloud to obtain the central axis of the pipe: ； In the formula, A parametric curve representing the central axis of the pipeline; and These represent the major and minor semi-axises after fitting the ellipse to the pipe cross-section; The parameter variables representing the parametric curve of the center axis; This represents the slope of the central axis.

6. The working method of a coal mine drainage pipeline installation robot according to claim 5, characterized in that: The expression for the dynamic cost function mentioned in step S2 is as follows: ； In the formula, , and These represent the axis deviation weight, obstacle avoidance weight, and climbing cost weight, respectively. Representing path points To the center axis of the pipeline The distance; Representing path points The cost of obstacles at the location; Representing path points Compared with the previous path point Path curvature between; The MOA* algorithm, with its dynamic cost function at its core, searches all path points to obtain the optimal path. Total path cost and the corresponding track movement parameter sequence, wherein the track movement parameters include the target climbing speed for each segment of the path. and steering angle .

7. The working method of a coal mine drainage pipeline installation robot according to claim 6, characterized in that: Step S3 specifically includes the following steps: S31. Calculate the total adsorption force of the target. : ； in, ； ； ； In the formula, Indicates the safety factor of the adsorption force; This represents the minimum threshold of total adsorption force; This represents the minimum total adsorption force against sliding constraints; This represents the minimum total adsorption force required to resist overturning. Indicates the coefficient of friction between the track and the installed inner wall; Indicates the inner radius of the installed pipe; This represents the robot's total weight. The horizontal lever arm representing the distance from the point of application of the adsorption force to the robot's center of gravity; S32. Based on the target total adsorption force, calculate the adsorption force required for a single set of electromagnetic adsorption points. and corresponding drive current and driving voltage : ； ； ； In the formula, This indicates the number of electromagnetic adsorption points that are active. This indicates the distance between the electromagnetic adsorption point and the air gap inside the pipe wall; Indicates the permeability of free space; This represents the effective area of ​​the adsorption surface of a single set of adsorption points; This indicates the number of turns of the electromagnetic coil at a single adsorption point; This represents the resistance of a single set of adsorption point coils; S33, Based on Optimal Path and track walking parameters, configure climbing speed Target speed of track motor And the rotational speeds on the inner and outer sides of the pipe bends: ； ； ； ； In the formula, Indicates the diameter of the track drive wheel; Indicates the actual speed of the drive motor; Indicates the transmission ratio of the harmonic reducer; This indicates the number of steps per revolution of the track drive motor; and These represent the speeds of the outer and inner tracks during turning, respectively. Indicates the center-to-center distance between the two tracks; Indicates the steering adjustment time; During the climb, a step-by-step strategy is implemented: activation of the first half of the adsorption points, transition switching, and activation of the second half of the adsorption points. The trigger distance for adsorption point switching is specified. Switching interval time and the adsorption force retention threshold during the switching process The calculation formula is as follows: ； ； ； In the formula, Indicates track pitch; Furthermore, during the climb, at each path point... Pause for 0.5 seconds, and use the attitude sensor results to correct the current adjustment amount based on the adsorption force of a single adsorption point. and attitude correction speed difference adjustment amount : ； ； in, ； ； In the formula, Indicates the attitude deviation value; This indicates the deviation value of the adsorption force; Indicates the robot's roll angle deviation; Indicates the robot's pitch angle deviation; This represents the actual total adsorption force; During the climb, ToF sensors and RGB-D cameras were used to collect real-time 3D distance data of the pipe's inner wall, as well as pipe images and depth data, to determine the climbing position. Simultaneously calculate the positioning error : ； ； In the formula, and These represent the ToF and RGB-D weights, respectively. Indicates the relative position based on the ToF sensor's own coordinate system; This indicates the pixel-depth correspondence based on the RGB-D camera's own coordinate system; This represents the three-dimensional world coordinates corresponding to the robot's actual climbing position after calibration. This represents the three-dimensional world coordinates corresponding to the target installation station that the robot is climbing; The final output is the corrected real-time climbing position. and positioning error and in positioning error At this time, the three-dimensional deviation data between the current climbing position and the target installation position is re-acquired, and the track axial displacement and differential steering angle are adjusted based on the three-dimensional deviation data until the requirements are met. .

8. The working method of a coal mine drainage pipeline installation robot according to claim 7, characterized in that: Step S4 specifically includes the following steps: S41, when the corrected real-time climbing position When the pipe reaches 50mm from the end face of the installed pipe, stop climbing and use the electromagnetic adsorption point for positioning. At this time, the three-jaw tensioning disc extends out of the installed pipe and drives the three-jaw tensioning disc to extend synchronously to tension the outer wall of the pipe to be installed until the actual tension force collected by the force sensor is equal to the target tension force. At this time, the axis of the pipe to be installed coincides with the axis of the installed pipe. S42. Activate the ADRC-fractional order fusion vibration suppression algorithm to construct a coupled vibration dynamic model of the three-jaw tensioning disc, the pipe to be installed, and the tracked motor: ； In the formula, This indicates the total mass of the three-jaw tensioner, the pipe to be installed, and the track motor; and These represent the system's damping coefficient and stiffness coefficient, respectively. , and These represent the vibration displacement, vibration velocity, and vibration acceleration along the pipeline axis, respectively. Indicates the disturbance force of the downhole environment; This indicates residual vibration interference force during track climbing; S43. Estimate the real-time total disturbance using a nonlinear observer based on a coupled vibration dynamics model. and will real-time total interference Decomposed into residual vibration interference force during track climbing and the disturbance force of the downhole environment Then, the disturbance force of the underground environment After elimination, the final total interference is obtained. : ； S44, Generating vibration suppression control quantity : ； in, ； In the formula, Indicates the basic control quantity; This represents the control gain coefficient; , , These represent the proportional, differential, and integral coefficients, respectively. Indicates vibration displacement deviation; Indicates the order of integration; This represents a fractional differential operator.

9. The working method of a coal mine drainage pipeline installation robot according to claim 8, characterized in that: In step S5, based on the vibration suppression control quantity Adjust the axial adjustment mechanism and the attitude adjustment mechanism to adjust the coaxiality between the pipe to be installed and the already installed pipe until the final coaxiality is obtained using video detection technology. .

10. The working method of a coal mine drainage pipeline installation robot according to claim 9, characterized in that: Step S6 specifically includes the following steps: S61. Perform semantic segmentation on the docking area, extract defect features, and calculate the gap width through curve fitting. : ； In the formula, and These represent the pixel coordinates on both sides of the gap to be detected; S62. Establish the SPH model for tube-to-tube interactions: ； In the formula, express; express; express; express; express; express; express; express; S63. For particles in the SPH model, their strain rate tensor The discrete form is: ； In the formula, Represents particles The quality; and They represent particles respectively and particles speed; and They represent particles respectively and particles The density; and They represent particles respectively and particles Position vector; Represents the sliding kernel function The gradient; S64. Obtain the particle by integrating the strain rate tensor. strain tensor : ； In the formula, Let represent the initial strain tensor, and ; S65. Calculation of stress tensor based on linear elastic constitutive equation : ； in, ； ； In the formula, and Both represent Lamé constants; Represents the trace of the strain tensor, and , , and They represent particles respectively Normal strain in the x, y, and z directions; Represents the Kronecker function; This indicates the elastic modulus of the pipe material; Indicates the Poisson's ratio of the pipe material; S66, stress tensor Converted to von Mises equivalent force : ； In the formula, , , These represent the normal stresses in the pipe along the x, y, and z directions, respectively. , , These represent the shear stresses of the pipe in the xy plane, yz plane, and zx plane, respectively. S67. Iterate through the equivalent stress of all pipe particles in the SPH model and take the maximum value as the maximum stress. : ； S68, Based on maximum stress Determine the risk level : ； In the formula, This indicates the allowable stress of the pipe material.

Images