Autonomous operation method and system of coal mine underground roadway repair robot

Abstract

The application provides an autonomous operation method and system of a coal mine underground roadway repair robot. The method comprises the following steps: positioning a deformation area based on three-dimensional point cloud data of a roadway collected by a plurality of laser radars, and planning a navigation path according to position information of the deformation area; detecting whether there is an exposed anchor rod in the deformation area through a first camera, and if there is, positioning the exposed anchor rod through the plurality of laser radars and the first camera, and controlling an auxiliary arm and shearing tongs to shear; controlling a work mechanical arm and a breaking hammer to break the deformation area, and identifying floating coal on a roadway floor through the plurality of laser radars, and controlling the work mechanical arm and a shovel to gradually clean the floating coal; controlling the repair robot to drive forward to clean a step distance, and detecting an anchor rod missing position through a second camera, and controlling an anchor rod drilling machine to set an anchor rod at the anchor rod missing position. The method realizes accurate control of the whole process of roadway repair operation, and improves the intelligence, efficiency and safety of roadway repair.

Description

Technical Field

[0001] This application relates to the field of underground roadway repair technology, and in particular to an autonomous operation method and system for a coal mine underground roadway repair robot. Background Technology

[0002] Currently, various underground operations in coal mines may cause roadway deformation, affecting the safety of underground work and requiring repair.

[0003] In related technologies, underground roadway repair mainly relies on manual operation. Most deformed roadways are repaired using manually operated pneumatic picks, excavators, and ordinary undercutters. However, this manual repair method involves extremely high labor intensity for workers, and the repair process lacks a systematic approach, resulting in long repair cycles, low repair efficiency, and high repair costs. Furthermore, during manual repair, workers must handle anchor bolts and fractured surrounding rock at close range in the deformation hazard zone, posing significant safety risks.

[0004] Therefore, how to control the tunnel repair equipment to carry out intelligent autonomous repair operations has become an urgent problem to be solved. Summary of the Invention

[0005] This application aims to at least partially address one of the technical problems in the related art.

[0006] Therefore, the first objective of this application is to propose an autonomous operation method for a coal mine underground roadway repair robot. This method achieves precise control of the entire roadway repair operation process through autonomous sensing using multiple detection devices on the repair robot, thereby improving the intelligence, efficiency, and safety of roadway repair. It solves the problems of high reliance on manual labor and low levels of automation and intelligence in some functions of roadway repair equipment.

[0007] The second objective of this application is to propose an autonomous operation system for a coal mine underground roadway repair robot.

[0008] The third objective of this application is to provide a non-transitory computer-readable storage medium.

[0009] To achieve the above objectives, the first aspect of this application is to propose an autonomous operation method for a coal mine underground roadway repair robot, wherein the cleaning robot includes: multiple lidar sensors, a first camera, a second camera, a working robotic arm, an auxiliary arm, a shearing shear, a hydraulic breaker, a bucket, and an onboard anchor drilling rig, and the method includes the following steps:

[0010] Based on the three-dimensional point cloud data of the tunnel collected by the multiple lidars, the deformed area to be repaired is located and the location information of the deformed area is obtained. A navigation path is planned according to the location information, and the repair robot is controlled to travel to the working position according to the navigation path.

[0011] The first camera detects whether there is an exposed anchor bolt in the deformation area. If there is an exposed anchor bolt, the multiple lidars and the first camera work together to locate the exposed anchor bolt, and the auxiliary arm and the shearing clamp are controlled to cut the exposed anchor bolt based on the anchor bolt location result.

[0012] The robotic arm and the breaker hammer are controlled to break up the deformed area without exposed anchor bolts, and the multiple lidars are used to identify floating coal on the roadway floor. Based on the floating coal identification results, the robotic arm and the bucket are controlled to gradually clean up the floating coal.

[0013] After the floating coal is cleaned up, the repair robot is controlled to move forward a preset cleaning distance, and the second camera is used to detect the missing anchor bolt positions on the coal wall surface. The on-board anchor bolt drilling rig is then controlled to drive anchor bolts into the missing anchor bolt positions.

[0014] Optionally, in one embodiment of this application, after locating the deformed area to be repaired and obtaining the location information of the deformed area, the method further includes: detecting the degree of deformation of the deformed area and determining whether the degree of deformation exceeds a safety threshold; if the degree of deformation exceeds the safety threshold, controlling the repair robot to suspend operation and report a warning message.

[0015] Optionally, in one embodiment of this application, the step of coordinating the positioning of the exposed anchor bolt by the plurality of lidars and the first camera includes: performing point cloud registration processing on the point cloud data of the exposed anchor bolt collected by the plurality of lidars and the first camera to obtain the three-dimensional pose information of the exposed anchor bolt; the step of controlling the auxiliary arm and the shearing clamp to cut the exposed anchor bolt based on the anchor bolt positioning result includes: planning the motion trajectory of the auxiliary arm based on the three-dimensional pose information using an inverse kinematics algorithm, and driving the shearing clamp to move to the target working point to cut the exposed anchor bolt according to the motion trajectory; if there are remaining exposed anchor bolts in the deformation area, repeating the anchor bolt positioning and autonomous cutting until there are no exposed anchor bolts in the deformation area.

[0016] Optionally, in one embodiment of this application, controlling the robotic arm and the hydraulic breaker to break the deformed area without exposed anchor bolts includes: based on a force-position hybrid control strategy, controlling the robotic arm to dynamically adjust the impact frequency and impact angle of the hydraulic breaker to drive the hydraulic breaker to perform the breaking operation.

[0017] Optionally, in one embodiment of this application, the step of identifying floating coal on the roadway floor using the multiple lidars includes: identifying the floating coal accumulation area and determining the floating coal position coordinates from the point cloud data of the roadway floor collected by the multiple lidars using a density clustering algorithm; the step of controlling the working robot arm and the bucket to gradually clear the floating coal based on the floating coal identification result includes: planning the movement trajectory of the working robot arm according to the floating coal position coordinates, and driving the bucket to scoop up the floating coal to the side conveyor belt according to the movement trajectory; after each round of scooping, evaluating the remaining coal amount in the floating coal accumulation area, and if the remaining coal amount does not meet the standard, re-determining the floating coal position coordinates and planning the movement trajectory of the current round of scooping task.

[0018] Optionally, in one embodiment of this application, the repair robot further includes a chassis and multiple detection devices. Controlling the repair robot to move forward by a preset cleaning step distance includes: combining feedback data from the multiple lidars and the multiple detection devices to correct the direction of travel of the repair robot in real time, wherein the repair robot is driven by the chassis; and avoiding obstacles encountered during travel by using dynamic avoidance algorithms.

[0019] Optionally, in one embodiment of this application, the step of detecting the missing anchor bolt location on the coal wall surface using the second camera includes: locating the position coordinates of the anchor bolt tray on the coal wall surface using the second camera, and determining the anchor bolt missing location coordinates based on the position coordinates of the anchor bolt tray and the location of the missing area; before controlling the airborne anchor bolt drilling rig to drive the anchor bolt to the missing location, the step further includes: adjusting the attitude of the airborne anchor bolt drilling rig based on the anchor bolt missing location coordinates.

[0020] To achieve the above objectives, a second aspect of this application also proposes an autonomous operating system for a coal mine underground roadway repair robot. The repair robot includes: multiple lidar sensors, a first camera, a second camera, a robotic arm, an auxiliary arm, a shearing shear, a hydraulic breaker, a bucket, and an onboard anchor drilling rig. The system includes the following modules:

[0021] The positioning module is used to locate the deformed area to be repaired and obtain the location information of the deformed area based on the three-dimensional point cloud data of the alleyway collected by the multiple lidars, and to plan a navigation path according to the location information to control the repair robot to travel to the working position.

[0022] The shearing module is used to detect whether there is an exposed anchor bolt in the deformation area through the first camera. If there is an exposed anchor bolt, the exposed anchor bolt is located by the multiple lidars and the first camera in coordination, and the auxiliary arm and the shearing clamp are controlled to cut the exposed anchor bolt based on the anchor bolt positioning result.

[0023] The crushing and shoveling module is used to control the working robotic arm and the breaker hammer to crush the deformed area without exposed anchor bolts, and to identify floating coal on the roadway floor through the multiple lidars. Based on the floating coal identification results, the module controls the working robotic arm and the bucket to gradually clean up the floating coal.

[0024] The anchoring module is used to control the repair robot to move forward a preset cleaning step distance after the floating coal is cleaned up, and to detect the missing anchor bolt positions on the coal wall surface through the second camera, and control the on-board anchor bolt drilling rig to install anchor bolts at the missing anchor bolt positions.

[0025] To implement the above embodiments, the third aspect of this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the autonomous operation method of the coal mine underground roadway repair robot in the first aspect.

[0026] The technical solution provided by the embodiments of this application brings at least the following beneficial effects: This application first locates the deformation area by collecting point cloud data through lidar, then cuts the exposed anchor bolts based on the recognition results of the front camera, then breaks up the deformation area, then cleans the loose coal based on the recognition results of lidar, and finally installs the anchor bolts based on the recognition results of the rear camera. Thus, this application, based on a fusion system of multiple lidar, multiple cameras, and IMU and other multi-sensor systems, can realize the autonomous and intelligent full-process operation of the roadway repair robot. Through the autonomous operation of the repair robot, on the one hand, the labor intensity of workers can be reduced, and the efficiency of roadway repair is improved through precise control of fully automated equipment, ensuring the accuracy of the repair operation. On the other hand, the unmanned repair method reduces the danger of roadway repair operations and ensures the safety of workers. The roadway cleaning robot of this application makes operational decisions entirely based on environmental perception information, improving roadway repair efficiency, enhancing the intelligence level of underground roadway repair equipment, and promoting the upgrading of underground operation safety. Therefore, this application improves the intelligence, efficiency, and safety of roadway repair.

[0027] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0028] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0029] Figure 1 A flowchart illustrating an autonomous operation method for a coal mine underground roadway repair robot proposed in this application embodiment;

[0030] Figure 2 This is a structural schematic of a tunnel repair robot proposed in an embodiment of this application;

[0031] Figure 3 This is a schematic diagram illustrating a specific tunnel repair process as proposed in an embodiment of this application;

[0032] Figure 4 This is a schematic diagram of the structure of an autonomous operation system for a coal mine underground roadway repair robot proposed in an embodiment of this application. Detailed Implementation

[0033] Embodiments of the present invention are described in detail below, examples of which are illustrated 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. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the present invention, and should not be construed as limiting the present invention.

[0034] It should be noted that the manual repair methods in related technologies lack a systematic approach, failing to simultaneously complete deformation monitoring, precise cleaning, and support reinforcement, resulting in low repair efficiency. While some related embodiments utilize automated equipment for repair, the following problems remain: First, the repair equipment has limited functionality and adaptability. Some automation systems only possess basic crushing and loading functions, still requiring manual operation to clean loose coal from the floor, and cannot achieve shearing of sidewall anchors, identification of deformed areas, or maintenance of surrounding rock stability. Furthermore, the equipment design is not adapted to various complex geological conditions underground, and its high cost makes large-scale application difficult. Second, the level of equipment automation is low. Current roadway cleaning machines primarily focus on floor cleaning, relying on manual assistance to handle sidewall anchor treatment and repair deformed areas. The equipment lacks autonomous perception and decision-making capabilities; key aspects such as deformation monitoring, target identification and positioning, path planning and obstacle avoidance, and maintenance of surrounding rock stability have not yet achieved intelligent operation, limiting operational accuracy and efficiency.

[0035] Therefore, this application proposes an autonomous operation method and system for a coal mine underground roadway repair robot, which breaks through the core technologies such as autonomous perception, precise operation and full-process closed-loop control, and can realize the unmanned, efficient and safe upgrade of roadway repair.

[0036] The following description, with reference to the accompanying drawings, illustrates an autonomous operation method and system for a coal mine underground roadway repair robot, as proposed in the embodiments of this application.

[0037] Figure 1 This is a flowchart illustrating an autonomous operation method for a coal mine underground roadway repair robot proposed in an embodiment of this application, as shown below. Figure 1 As shown, the method includes the following steps:

[0038] Step S101: Based on the three-dimensional point cloud data of the tunnel collected by multiple lidars, the deformed area to be repaired is located and the location information of the deformed area is obtained. Based on the location information, a navigation path is planned and the repair robot is controlled to travel to the working position according to the navigation path.

[0039] Specifically, when it is necessary to repair the deformed areas of underground roadways, the roadway repair robot is first activated. Multiple lidar sensors on the robot dynamically scan the roadway to detect the deformed areas. The repair robot described in this application includes at least the following modules: multiple lidar sensors, a first camera, a second camera, a robotic arm, an auxiliary arm, a shearing pliers, a hydraulic breaker, a bucket, and an onboard anchor drilling rig.

[0040] In one embodiment of this application, such as Figure 2 As shown, the tunnel repair robot targeted in this application includes: a first lidar 1, a second lidar 2, a first camera 3, a second camera 4, a working robotic arm 5, an auxiliary arm 6, a shearing shear 7, a hydraulic breaker 8, a bucket 9, an onboard anchor drilling rig 10, a personnel platform 11, a chassis 12, and a power system 13. This embodiment uses two lidars and two cameras as an example. In practical applications, the number of lidars and cameras can be determined as needed, and each lidar and camera can be installed in a convenient scanning position. For example, Figure 2 As shown, the first camera 3 is set in front of the robot to capture the area in front, and the second camera 4 is set in the rear to identify the area behind the robot.

[0041] To more clearly describe the specific process of the roadway repair robot in this application for roadway repair, the following is an example of the roadway repair robot proposed in the embodiments of this application. Figure 3 The work process shown is illustrated by example. Specifically, as... Figure 3 As shown, when the repair robot operates autonomously, it first powers on and starts up. After each module within the robot is powered on, it initializes synchronously, automatically completing hardware self-tests and sensor calibrations. This includes initializing the LiDAR, vision camera, and robotic arm drive modules, with each subsystem entering standby mode.

[0042] Furthermore, tunnel scanning modeling and environmental perception are performed. Multiple LiDAR sensors mounted on the robot scan the tunnel ahead in real time, generating 3D point cloud data with centimeter-level accuracy. The collected point cloud data is then analyzed to construct a tunnel point cloud map.

[0043] One possible approach is to first combine simultaneous localization and mapping (SLAM) algorithms to build a real-time geometric model of the tunnel. Then, feedback data from the inertial measurement unit (IMU) pre-installed on the robot is simultaneously fused to correct robot pose offsets during point cloud acquisition and data computation, establishing a dynamically updated global map. This ensures that the tunnel map is consistent with the actual environment, providing an environmental benchmark for subsequent operations.

[0044] Specifically, in this embodiment, the LiDAR continuously scans at different angles to obtain scan frames at different times. Then, the SLAM algorithm is used to match and fuse the scan frames with different timestamps, eliminating motion distortion and constructing a high-precision global point cloud map of the tunnel. Next, the robot's pose is corrected by repairing feedback data from multiple detection devices within the robot, including dynamically correcting the robot's pose offset based on data from its odometry and IMU. Finally, the tunnel point cloud map is recalculated and updated based on the pose correction results, generating an accurate global tunnel map.

[0045] Furthermore, the deformation areas are extracted and located. Based on the obtained global map data of the tunnel, the maximum deformation area of ​​the tunnel wall (i.e., the local collapse area) is identified using a planar extraction algorithm, and the location information of the deformation area, such as the boundary coordinates of the deformation area, is extracted.

[0046] One possible implementation is to first segment multiple planar regions constituting the roadway from the preprocessed roadway point cloud model using the Random Sample Consensus (RANSAC) algorithm. Then, a clustering algorithm is used to cluster the geometry of these planar regions, and the largest cluster is selected from the resulting clusters. The plane corresponding to the largest cluster is then taken as the region with the largest deformation. Next, the contour and boundary coordinates of the detected deformation region are extracted. For example, the deformation region can be projected onto a two-dimensional plane, and a convex hull fitting is performed on the point set on the two-dimensional plane to obtain multiple convex polygons. The smallest convex polygon is taken as the contour of the deformation region. Then, the coordinates of the contour in the two-dimensional plane and three-dimensional space are analyzed to obtain the boundary coordinates of the deformation region.

[0047] Furthermore, based on the coordinate location information of the deformed area, the navigation path of the repair robot can be planned, and the repair robot can be controlled to travel along the navigation path to the working position for subsequent repair work.

[0048] One possible approach is to determine the coordinates of the repair work area based on the coordinates of the deformed region, and use these coordinates as the target coordinates for the robot's navigation endpoint. Then, based on the robot's starting and target coordinates, and considering factors such as the underground tunnel's geographical information and equipment information, real-time path planning is performed. Following the planned optimal path, the repair robot is navigated to the target coordinates.

[0049] Based on the above embodiments, in order to ensure the safety of the repair robot during repair operations, it can also be determined whether the deformed area of ​​the roadway supports autonomous robot operation. In one embodiment of this application, after locating the deformed area to be repaired and obtaining the location information of the deformed area, the method further includes: detecting the degree of deformation of the deformed area and determining whether the degree of deformation exceeds a safety threshold; if the degree of deformation exceeds the safety threshold, controlling the repair robot to suspend operation and reporting a warning message.

[0050] Specifically, in this embodiment, the degree of deformation in the deformed area can be detected by various methods, such as calculating the difference between the real-time boundary coordinates of the deformed area and the boundary coordinates of the area before deformation. If the degree of deformation in the deformed area exceeds a preset safety threshold, it indicates that the roadway currently poses a significant safety hazard and is not suitable for repair work. Subsequently, the repair robot is controlled to pause its work, and a warning message is sent to the remote control center via the robot's communication system, indicating that the roadway is currently dangerous, so that relevant safety measures can be taken in a timely manner.

[0051] In step S102, the first camera detects whether there is an exposed anchor bolt within the deformation area. If there is an exposed anchor bolt, multiple lidars and the first camera work together to locate the exposed anchor bolt, and the auxiliary arm and shearing clamps are controlled to cut the exposed anchor bolt based on the anchor bolt location result.

[0052] Specifically, when the repair robot performs the repair work, it first performs exposed anchor detection and autonomous shearing. The first camera detects whether there are exposed anchors in the deformation area. If there are no exposed anchors, it directly proceeds to step S103. If there are exposed anchors, multiple LiDARs and the first camera work together to locate the exposed anchors.

[0053] In one embodiment of this application, the exposed anchor bolt is located collaboratively by multiple lidars and a first camera, including: performing point cloud registration processing on the point cloud data of the exposed anchor bolt collected by multiple lidars and the first camera to obtain the three-dimensional pose information of the exposed anchor bolt.

[0054] Specifically, point cloud data of the deformation area is collected by multiple lidar sensors and a first camera 3. Since the acquisition angles and point cloud data processing procedures of the multiple lidar sensors and the first camera 3 differ, this embodiment uses point cloud registration technology to process the data collected by the multiple detection devices, thereby obtaining accurate three-dimensional pose information of the exposed anchor bolt. The obtained three-dimensional pose information includes the position coordinates and tilt angle of the exposed anchor bolt.

[0055] As one possible approach, when acquiring 3D pose data using point cloud registration technology, preprocessing of different point cloud data is first performed to improve the efficiency and accuracy of registration. This includes downsampling, noise removal, and feature point extraction. Then, coarse registration is performed, finding an approximate rotation and translation matrix between the two point clouds acquired by multiple LiDARs and the first camera 3, roughly aligning their relative positions. Fine registration is then performed, further optimizing the rotation and translation matrix based on the coarse registration to achieve more precise alignment. For example, the ICP (Iterative Closest Point) algorithm can be used. Finally, based on the rotation matrix and translation vector obtained from point cloud registration, the 3D pose information of the exposed anchor bolt can be obtained.

[0056] Furthermore, in this embodiment, controlling the auxiliary arm and shearing clamp to cut exposed anchor bolts based on anchor bolt positioning results includes: planning the motion trajectory of the auxiliary arm based on three-dimensional pose information using an inverse kinematics algorithm, and driving the shearing clamp to move to the target working point to cut the exposed anchor bolts according to the motion trajectory; if there are remaining exposed anchor bolts in the deformation area, repeating anchor bolt positioning and autonomous shearing until there are no exposed anchor bolts in the deformation area.

[0057] Specifically, such as Figure 3 As shown, in this embodiment, the motion trajectory of the auxiliary arm 6 is autonomously planned based on the three-dimensional pose information of the anchor rod. Specifically, the inverse kinematics algorithm is used to plan a collision-free trajectory, driving the current attachment (i.e., the shearing clamp 7) to move precisely to the target point. The final target coordinates of the auxiliary arm 6 after it stops running can be determined based on the coordinates of the exposed anchor rod, so as to move the shearing clamp 7 to a suitable position to shear the exposed anchor rod.

[0058] As one possible implementation, when planning a collision-free trajectory using inverse kinematics algorithms, inverse kinematics is first solved to calculate the joint angles of the robot's auxiliary arm 6, ensuring the end effector reaches the target coordinates and corresponding posture. Then, trajectory planning is performed. Based on the solved inverse kinematics, a collision-free trajectory from the initial position to the target coordinates is planned. This process can employ common trajectory planning methods, such as time-parameterized trajectory planning or convex optimization-based trajectory planning. To ensure the trajectory is collision-free, collision detection and optimization are finally required. For example, collision detection involves real-time monitoring of whether points on the trajectory collide with obstacles in the environment during the planning process. Spatial partitioning methods (such as octrees and VDBs) can be used to accelerate collision detection. For path optimization, convex optimization methods (such as GCS, Graphs of Convex Sets) can be used to optimize the trajectory. The GCS method generates a collision-free trajectory by dividing the configuration space into convex safe regions and finding the shortest path between these regions.

[0059] Furthermore, after the shearing clamp 7 has finished shearing the current exposed anchor bolt, a second inspection is performed on the remaining exposed anchor bolt. The anchor bolt positioning, autonomous trajectory planning and autonomous shearing in the above embodiment are executed repeatedly until there are no exposed anchor bolts in the deformation area, at which point subsequent cleaning operations can be carried out.

[0060] Step S103: Control the robotic arm and breaker to break the deformed area without exposed anchor bolts, and use multiple lidar to identify floating coal on the roadway floor. Based on the floating coal identification results, control the robotic arm and bucket to gradually clear the floating coal.

[0061] Specifically, after the residual anchor bolts are cleaned up, a crushing and brushing operation is carried out. In this operation, the attachment is switched to a hydraulic breaker 8, and then the deformed area is crushed.

[0062] In one embodiment of this application, controlling the robotic arm and the breaker to break up a deformed area without exposed anchor bolts includes: based on a force-position hybrid control strategy, controlling the robotic arm to dynamically adjust the impact frequency and impact angle of the breaker to drive the breaker to perform the breaking operation.

[0063] Specifically, in this embodiment, the robotic arm dynamically adjusts the impact frequency and angle based on a force-position hybrid control strategy to avoid excessive damage to the tunnel structure. As one possible implementation, the impact frequency can be adjusted through mechanical and hydraulic adjustments of the robotic arm 5. For example, the piston stroke can be changed by adjusting screws on the top or side of the hydraulic breaker's cylinder, thereby adjusting the impact frequency. The working pressure and flow rate of the hydraulic system can also be adjusted to change the impact frequency of the breaker. The impact angle can be adjusted through the installation position of the breaker and the angle of the robotic arm 5. During the above adjustments, precise control of the impact frequency and angle can be achieved by adjusting the force-position hybrid control parameters. For example, the inertia parameter, stiffness parameter, and damping parameter, among other force-position hybrid control parameters, can be adjusted individually.

[0064] Further, the loose coal on the roadway floor is detected and loaded / unloaded. First, multiple lidar sensors scan the roadway floor, and the location information of the loose coal is identified based on the collected floor point cloud data. Then, the detected loose coal is shoveled and loaded into other areas using the bucket 9.

[0065] In one embodiment of this application, identifying floating coal on the roadway floor using multiple lidars includes: identifying the floating coal accumulation area and determining the location coordinates of the floating coal from the point cloud data of the roadway floor collected by multiple lidars using a density clustering algorithm.

[0066] Specifically, this embodiment identifies floating coal accumulation areas using density clustering algorithms. For example, algorithms such as DBSCAN, OPTICS, or HDBSCAN can be selected. The parameters of the chosen algorithm are adjusted based on the density distribution of point cloud data collected by multiple lidar sensors. The adjusted density clustering algorithm is then run to perform clustering processing, identifying floating coal accumulation areas in the point cloud. The clustering results can be analyzed to extract the coordinate information of the clustered floating coal accumulation areas. Furthermore, based on the clustering results of the floating coal accumulation areas, a heat map of the floating coal distribution can be output to clearly and intuitively display the floating coal located at different positions on the bottom plate.

[0067] Furthermore, this embodiment controls the robotic arm and bucket to gradually clear the floating coal based on the floating coal identification results, including: planning the movement trajectory of the robotic arm according to the location coordinates of the floating coal, and driving the bucket to scoop up the floating coal to the side conveyor belt according to the movement trajectory; after each round of scooping, assessing the amount of remaining coal in the floating coal accumulation area, and if the amount of remaining coal does not meet the standard, re-determining the location coordinates of the floating coal and planning the movement trajectory of the current round of scooping task.

[0068] Specifically, in this embodiment, the attachment of the robotic arm 5 is first switched to a bucket 9. Based on parameters such as the volume and coordinates of the coal pile to be cleared, the shoveling trajectory of the robotic arm 5 is planned. A step-by-step shoveling strategy is adopted to shovel the loose coal onto the conveyor belt on the side of the robot. For example... Figure 3 As shown, after each loading operation, the remaining amount of floating coal is re-scanned and evaluated according to the floating coal detection method described in the above embodiment. If the remaining amount of floating coal does not meet the standard, for example, if the remaining amount of floating coal is greater than the allowable residual coal amount, it indicates that the floating coal has not been completely cleared, and the next round of loading task is carried out. The working path for this round of floating coal shoveling is then re-planned according to the motion trajectory planning method of the robotic arm 5 described in the above embodiment.

[0069] Therefore, this embodiment, by employing a distributed shovel strategy for loose coal, can avoid the impact of changes in the position of loose coal caused by each shovel loading. Each round of shoveling task uses the most reasonable movement trajectory to shovel loose coal, thereby further improving the working efficiency of the repair robot.

[0070] Step S104: After the floating coal is cleaned up, control the repair robot to move forward by a preset cleaning step distance, and use the second camera to detect the missing anchor bolt positions on the coal wall surface, and control the on-board anchor bolt drill to drive anchor bolts to the missing anchor bolt positions.

[0071] Specifically, after the floating coal is cleaned up, the repair robot is first controlled to move forward, and the forward movement is controlled by a preset step distance.

[0072] In one embodiment of this application, the repair robot further includes a chassis and multiple detection devices. Controlling the repair robot to move forward by a preset cleaning step distance includes: combining feedback data from multiple lidars and multiple detection devices to correct the direction of travel of the repair robot in real time, wherein the repair robot is driven by the chassis; and avoiding obstacles encountered during travel by using dynamic avoidance algorithms.

[0073] Specifically, this embodiment utilizes a tracked chassis 12 to propel the repair robot autonomously forward a preset cleaning step distance (e.g., 0.5 meters). This step distance can be determined based on the location of the loose coal and the position of the anchor bolt area on the coal face; the specific step length is calculated according to actual conditions. During forward movement, to avoid deviation from the travel direction and collisions with unexpected obstacles, this embodiment integrates data from multiple lidar sensors and pre-installed visual odometry devices on the robot into the navigation system, correcting the travel direction in real time to ensure consistency with the planned direction. When an obstacle is encountered during travel, a dynamic obstacle avoidance algorithm is triggered, allowing the robot to detour and return to the original planned path.

[0074] One possible approach is to first fuse the data from LiDAR and visual odometry to overcome the limitations of a single sensor and improve the accuracy and robustness of robot localization. Specifically, various coupling methods, such as tightly coupled fusion and loosely coupled fusion, can be used to fuse the corresponding robot pose estimates from LiDAR and visual odometry to obtain the robot's global pose. Then, based on the fused pose, the robot's forward path is planned to ensure it travels along the predetermined trajectory. During travel, control commands are generated in real time based on the path planning results to adjust the robot's forward direction.

[0075] When encountering obstacles during navigation, the robot first predicts the obstacle trajectory and then uses algorithms such as Dynamic Windowing (DWA) to plan an obstacle avoidance path. Next, an obstacle avoidance strategy is formulated, including speed adjustments and maintaining a safe distance, to ensure the robot can avoid obstacles. Finally, by updating feedback instructions in real time, the robot returns to its original path after confirming it has bypassed the obstacle.

[0076] Furthermore, after the repair robot completes its cleaning distance, automatic anchor bolt replacement and support repair are performed. A second camera 4 located behind the repair robot scans the coal face at the current position, identifies the missing anchor bolt locations in the anchor bolt tray, and then controls the onboard anchor bolt drill to install anchor bolts at the missing locations.

[0077] In one embodiment of this application, detecting the missing anchor bolt location on the coal face surface using a second camera includes: locating the position coordinates of the anchor bolt tray on the coal face surface using the second camera, and determining the coordinates of the missing anchor bolt location based on the position coordinates of the anchor bolt tray and the location of the missing area; before controlling the airborne anchor bolt drilling rig to drive the anchor bolt to the missing location, the method further includes: adjusting the attitude of the airborne anchor bolt drilling rig based on the coordinates of the missing anchor bolt location.

[0078] Specifically, in this embodiment, the data scanned by the second camera 4 is analyzed first. Using various target detection algorithms, such as the YOLO (You Only Look Once) model, the position of the anchor bolt tray on the coal face and the coordinates of various parts within the tray are determined. Then, the coordinates of the missing anchor bolt locations are determined, thus obtaining the coordinates of the missing anchor bolt positions. Next, the onboard anchor bolt drilling rig 10 is controlled to automatically adjust its posture based on the coordinates of the missing anchor bolt positions. For example, the position and angle of the onboard anchor bolt drilling rig 10 are adjusted to align with the coordinates of the missing anchor bolt positions. Then, the "one-click anchor bolt installation" program is initiated to complete the secondary reinforcement support. If there are multiple missing anchor bolt locations, the above process can be repeated for multiple rounds of anchor bolt installation.

[0079] Thus, this application completes a single autonomous operation cycle for the repair robot, and the tunnel cleaning robot can also wait for continuous operation instructions to achieve closed-loop autonomous operation throughout the entire process, and autonomously repair different deformed areas in different tunnels.

[0080] In summary, the tunnel deformation detection method based on a coal mine underground cleaning robot in this application first uses a lidar on the cleaning robot to dynamically scan the tunnel and collect 3D point cloud data of the tunnel in real time. Then, point cloud data from different time frames are registered to construct a tunnel point cloud map, and the robot's pose is corrected and the tunnel point cloud map is updated to generate a tunnel point cloud model. Next, multiple planar regions are segmented from the preprocessed tunnel point cloud model, and the geometric structures in these regions are clustered to obtain deformable regions. Finally, the contours of the deformable regions are extracted to determine multiple state information of the deformable regions. Therefore, this method utilizes the autonomous perception and calculation of the tunnel cleaning robot before operation, and performs deformation detection by scanning the tunnel's perception point cloud data. It can automatically detect deformable regions in the tunnel, obtain multiple accurate information about the deformable regions, accurately quantify the degree of deformation, and improve the accuracy and real-time performance of deformable region detection. Furthermore, this method requires no manual intervention throughout the entire detection process, reducing the workload of staff, improving the convenience of deformation area detection, ensuring staff safety, reducing the complexity of deformation area detection, and improving the efficiency of deformation area detection. Moreover, the automatically detected deformation information can be used for subsequent autonomous control of the robotic arm, realizing a closed-loop operation from perception to decision-making to execution, which is beneficial to improving the operational efficiency and intelligent control level of the roadway cleaning robot.

[0081] To achieve the above embodiments, this application also proposes an autonomous operation system for a coal mine underground roadway repair robot. Figure 4 This is a schematic diagram of the structure of an autonomous operating system for a coal mine underground roadway repair robot proposed in an embodiment of this application, as shown below. Figure 4 As shown, the system includes: a positioning module 100, a shearing module 200, a crushing and shoveling module 300, and an anchoring module 400.

[0082] The positioning module 100 is used to locate the deformed area to be repaired based on the three-dimensional point cloud data of the alleyway collected by multiple lidars, obtain the location information of the deformed area, plan a navigation path according to the location information, and control the repair robot to travel to the working position.

[0083] The shearing module 200 is used to detect whether there is an exposed anchor bolt in the deformation area through the first camera. If there is an exposed anchor bolt, the exposed anchor bolt is located by multiple lidars and the first camera in coordination, and the auxiliary arm and shearing clamp are controlled to cut the exposed anchor bolt based on the anchor bolt positioning result.

[0084] The crushing and shoveling module 300 is used to control the robotic arm and breaker hammer to crush deformed areas without exposed anchor bolts, and to identify floating coal on the roadway floor through multiple lidar sensors. Based on the floating coal identification results, the robotic arm and bucket are controlled to gradually clear the floating coal.

[0085] The anchoring module 400 is used to control the repair robot to move forward a preset cleaning distance after the floating coal is cleaned, and to detect the missing anchor bolt positions on the coal wall surface through a second camera, and control the on-board anchor bolt drilling rig to install anchor bolts at the missing anchor bolt positions.

[0086] Optionally, in one embodiment of this application, the system further includes an early warning module, which is specifically used to: detect the degree of deformation in the deformed area and determine whether the degree of deformation exceeds a safety threshold; if the degree of deformation exceeds the safety threshold, control the repair robot to suspend operation and report the early warning information.

[0087] Optionally, in one embodiment of this application, the shearing module 200 is specifically used for: performing point cloud registration processing on the point cloud data of exposed anchor bolts collected by multiple lidars and a first camera to obtain the three-dimensional pose information of the exposed anchor bolts; based on the three-dimensional pose information, planning the motion trajectory of the auxiliary arm through an inverse kinematics algorithm, and driving the shearing clamp to move to the target working point to shear the exposed anchor bolts according to the motion trajectory; if there are remaining exposed anchor bolts in the deformation area, repeating the anchor bolt positioning and autonomous shearing until there are no exposed anchor bolts in the deformation area.

[0088] Optionally, in one embodiment of this application, the crushing and shoveling module 300 is specifically used to: control the working robot arm to dynamically adjust the impact frequency and impact angle of the breaker hammer based on a force-position hybrid control strategy, so as to drive the breaker hammer to perform crushing operations.

[0089] Optionally, in one embodiment of this application, the crushing and shoveling module 300 is specifically used to: identify the floating coal accumulation area and determine the floating coal position coordinates from the point cloud data of the roadway floor collected by multiple lidars using a density clustering algorithm; plan the motion trajectory of the working robot arm according to the floating coal position coordinates, and drive the bucket to shovel the floating coal to the side conveyor according to the motion trajectory; after each round of shoveling, evaluate the remaining amount of coal in the floating coal accumulation area, and if the remaining amount of coal does not meet the standard, re-determine the floating coal position coordinates and plan the motion trajectory of the current round of shoveling task.

[0090] Optionally, in one embodiment of this application, the anchoring module 400 is specifically used to: combine feedback data from multiple lidars and multiple detection devices to correct the travel direction of the repair robot in real time, wherein the repair robot is driven by the chassis; and avoid obstacles encountered during travel through a dynamic avoidance algorithm.

[0091] Optionally, in one embodiment of this application, the anchoring module 400 is specifically used for: locating the position coordinates of the anchor tray on the coal wall surface using a second camera, and determining the coordinates of the missing anchor position based on the position coordinates of the anchor tray and the location of the missing area; and adjusting the attitude of the onboard anchor drilling rig based on the coordinates of the missing anchor position.

[0092] It should be noted that the explanation of the aforementioned embodiment of the autonomous operation method of the coal mine underground roadway repair robot also applies to the system of this embodiment, and will not be repeated here.

[0093] In summary, the autonomous operation system of the coal mine underground roadway repair robot in this application embodiment utilizes automatically detected deformation information for subsequent autonomous control of the robotic arm, realizing a closed-loop full-process operation from perception to decision-making to execution, which is conducive to improving the operation efficiency and intelligent control level of the roadway cleaning robot.

[0094] To implement the above embodiments, this application also proposes a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the autonomous operation method of the coal mine underground roadway repair robot as described in any one of the first aspects of the embodiments above.

[0095] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0096] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0097] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0098] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0099] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0100] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0101] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0102] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. An autonomous operation method for a coal mine underground roadway repair robot, characterized in that, The repair robot includes: multiple lidar sensors, a first camera, a second camera, a working robotic arm, an auxiliary arm, a shearing pliers, a hydraulic breaker, a bucket, and an onboard anchor drilling rig. The method includes the following steps: Based on the three-dimensional point cloud data of the tunnel collected by the multiple lidars, the deformed area to be repaired is located and the location information of the deformed area is obtained. A navigation path is planned according to the location information, and the repair robot is controlled to travel to the working position according to the navigation path. The first camera detects whether there is an exposed anchor bolt in the deformation area. If there is an exposed anchor bolt, the multiple lidars and the first camera work together to locate the exposed anchor bolt, and the auxiliary arm and the shearing clamp are controlled to cut the exposed anchor bolt based on the anchor bolt location result. The robotic arm and the breaker hammer are controlled to break up the deformed area without exposed anchor bolts, and the multiple lidars are used to identify floating coal on the roadway floor. Based on the floating coal identification results, the robotic arm and the bucket are controlled to gradually clean up the floating coal. After the floating coal is cleaned up, the repair robot is controlled to move forward a preset cleaning distance, and the second camera is used to detect the missing anchor bolt positions on the coal wall surface. The on-board anchor bolt drilling rig is then controlled to drive anchor bolts into the missing anchor bolt positions.

2. The method according to claim 1, characterized in that, After locating the deformed area to be repaired and obtaining the location information of the deformed area, the method further includes: The degree of deformation in the deformed region is detected, and it is determined whether the degree of deformation exceeds a safety threshold. If the degree of deformation exceeds the safety threshold, the repair robot is controlled to suspend operation and report an early warning.

3. The method according to claim 1, characterized in that, The method of locating the exposed anchor bolt in coordination with the multiple lidar sensors and the first camera includes: Point cloud registration processing is performed on the point cloud data of the exposed anchor bolts collected by the multiple lidars and the first camera to obtain the three-dimensional pose information of the exposed anchor bolts; The method of controlling the auxiliary arm and the shearing pliers to cut the exposed anchor bolt based on the anchor bolt positioning result includes: Based on the three-dimensional pose information, the motion trajectory of the auxiliary arm is planned by the inverse kinematics algorithm, and the auxiliary arm drives the shearing clamp to move to the target working point to cut the exposed anchor rod according to the motion trajectory. If there are still exposed anchor bolts in the deformation area, repeat the anchor bolt positioning and autonomous shearing until there are no exposed anchor bolts in the deformation area.

4. The method according to claim 1, characterized in that, The control of the robotic arm and the breaker hammer to break the deformed area without exposed anchor bolts includes: Based on the force-position hybrid control strategy, the working robot arm is controlled to dynamically adjust the impact frequency and impact angle of the breaker hammer in order to drive the breaker hammer to perform breaker operations.

5. The method according to claim 1, characterized in that, The process of identifying floating coal on the roadway floor using the multiple lidar sensors includes: The density clustering algorithm is used to identify the coal accumulation area and determine the location coordinates of the coal from the point cloud data of the roadway floor collected by the multiple lidars. The step of controlling the robotic arm and bucket to gradually remove floating coal based on the floating coal identification results includes: The movement trajectory of the robotic arm is planned according to the coordinates of the floating coal position, and the robotic arm drives the bucket to scoop up the floating coal to the side conveyor belt according to the movement trajectory. After each round of loading is completed, the amount of remaining coal in the floating coal accumulation area is assessed. If the amount of remaining coal does not meet the standard, the coordinates of the floating coal position are re-determined and the movement trajectory of the current round of loading is planned.

6. The method according to claim 1, characterized in that, The repair robot also includes a chassis and multiple detection devices. Controlling the repair robot to move forward a preset cleaning step distance includes: By combining the feedback data from the multiple lidars and the multiple detection devices, the travel direction of the repair robot is corrected in real time, wherein the repair robot is driven by the chassis. The dynamic avoidance algorithm avoids obstacles encountered during driving.

7. The method according to claim 1, characterized in that, The step of detecting the location of missing anchor bolts on the coal face using the second camera includes: The second camera is used to locate the position coordinates of the anchor tray on the coal wall surface, and the position coordinates of the missing anchor are determined based on the position coordinates of the anchor tray and the location of the missing area. Before controlling the airborne anchor drilling rig to drive an anchor bolt to the location where the anchor bolt is missing, the method further includes: The attitude of the airborne anchor drilling rig is adjusted according to the coordinates of the missing anchor bolt location.

8. An autonomous operating system for a coal mine underground roadway repair robot, characterized in that, The repair robot includes: multiple lidar sensors, a first camera, a second camera, a working robotic arm, an auxiliary arm, a shearing pliers, a hydraulic breaker, a bucket, and an onboard anchor drilling rig. The system includes the following modules: The positioning module is used to locate the deformed area to be repaired and obtain the location information of the deformed area based on the three-dimensional point cloud data of the alleyway collected by the multiple lidars, and to plan a navigation path according to the location information to control the repair robot to travel to the working position. The shearing module is used to detect whether there is an exposed anchor bolt in the deformation area through the first camera. If there is an exposed anchor bolt, the exposed anchor bolt is located by the multiple lidars and the first camera in coordination, and the auxiliary arm and the shearing clamp are controlled to cut the exposed anchor bolt based on the anchor bolt positioning result. The crushing and shoveling module is used to control the working robotic arm and the breaker hammer to crush the deformed area without exposed anchor bolts, and to identify floating coal on the roadway floor through the multiple lidars. Based on the floating coal identification results, the module controls the working robotic arm and the bucket to gradually clean up the floating coal. The anchoring module is used to control the repair robot to move forward a preset cleaning step distance after the floating coal is cleaned up, and to detect the missing anchor bolt positions on the coal wall surface through the second camera, and control the on-board anchor bolt drilling rig to install anchor bolts at the missing anchor bolt positions.

9. The system according to claim 8, characterized in that, It also includes an early warning module, which is specifically used for: The degree of deformation in the deformed region is detected, and it is determined whether the degree of deformation exceeds a safety threshold. If the safety threshold is exceeded, the repair robot is controlled to suspend its operation and report an early warning.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the autonomous operation method of the coal mine underground roadway repair robot as described in any one of claims 1-7.

Images