An autonomous operation control method and system of a mine intelligent pressure relief machine

Abstract

The application discloses a kind of self-operation control method and system of mine intelligent pressure relief machine, and the self-operation control method includes: obtaining the real-time image data of roadway surrounding rock, the image data is inferred based on YOLOv11 model, high-risk area and corresponding spatial position information are identified;Fusion laser radar point cloud data and visual image data, construct the three-dimensional semantic map of roadway, and plan the autonomous walking path based on the spatial position information of high-risk area;Control pressure relief machine moves to target operation point along the autonomous walking path, controls drill rod rotating mechanism and telescopic mechanism to execute drilling operation according to preset pressure relief parameter;When drilling to preset depth, automatically trigger reaming instruction, control umbrella type hydraulic reaming mechanism to open radially, and execute variable-diameter reaming operation;Obtain the ultrasonic imaging data in borehole and surrounding rock microseismic monitoring data, invert hole stress distribution state, and generate pressure relief effect evaluation result.

Description

Technical Field

[0001] This invention belongs to the field of intelligent mining machinery technology, and relates to an autonomous operation control method and system for an intelligent mining depressurization machine. Background Technology

[0002] Rockburst is a common dynamic disaster in deep mining of coal and non-coal mines. Its essence is a stress-rock mass coupling instability phenomenon induced by high stress concentration in the surrounding rock. Drilling stress relief technology, which involves drilling in the high-stress zone of the roadway sidewalls to disrupt the surrounding rock structure and promote stress transfer to deeper layers of the rock, is one of the effective means of preventing rockburst. This technology has the advantages of convenient construction and low cost, and has been widely used in mining engineering.

[0003] In practical engineering, it has been found that there are significant differences in the required borehole diameter between shallow and deep surrounding rock: if large-diameter boreholes are used in shallow surrounding rock, excessive stress relief can easily damage the roadway support structure; while in deep surrounding rock, sufficiently large borehole diameters are required to effectively transfer high stress to the original rock stress zone. Therefore, variable-diameter stress relief technology has emerged, which involves drilling conventional-diameter boreholes in the shallow surrounding rock section, and then enlarging the borehole after drilling to the deeper sections, thus balancing the safety of shallow support with the effectiveness of deep stress relief.

[0004] Currently, variable diameter depressurization operations mainly rely on manual operation or semi-automated equipment, which presents the following technical problems: First, determining the operating parameters is difficult. The selection of parameters such as drilling depth, hole enlargement timing, and hole enlargement diameter lacks real-time data support and often relies on engineering experience or preliminary theoretical calculations. It is difficult to dynamically adjust according to the actual state of the surrounding rock, which can easily lead to insufficient or excessive pressure relief.

[0005] Second, the operation process lacks autonomy. Existing equipment typically requires manual identification of high-risk areas, manual positioning of work points, and manual control of drill rod posture and hole enlargement actions. Operators must work in close proximity to high-stress areas, facing the safety risk of sudden rock bursts.

[0006] Third, there is a lack of a closed-loop feedback mechanism. After completing the drilling operation, the existing equipment cannot evaluate the pressure relief effect in real time, nor can it optimize subsequent operation parameters based on the evaluation results, resulting in inconsistent pressure relief quality in different sections of the same roadway.

[0007] Fourth, insufficient environmental adaptability. The underground environment is complex, with risks of dust, high humidity, vibration interference, and explosive gases. Existing equipment has weak visual perception, autonomous navigation, and safety interlocking capabilities, making it difficult to achieve stable and reliable autonomous operation under harsh conditions.

[0008] With the advancement of smart mine construction, mines are placing higher demands on the intelligence and autonomy of their equipment. Currently, no intelligent pressure relief equipment for mining can achieve closed-loop control throughout the entire process of "environmental perception—path planning—autonomous drilling—diameter-changing borehole enlargement—effect evaluation." Therefore, there is an urgent need for an intelligent control method and system capable of autonomously identifying high-risk areas, autonomously planning paths, precisely controlling diameter-changing operations, and possessing the ability to evaluate pressure relief effects and optimize parameters. This would replace manual operation and improve the safety, accuracy, and intelligence level of pressure relief operations.

[0009] Therefore, there is an urgent need to design an autonomous operation control method and system for intelligent depressurization machines used in mines to solve the technical problems existing in the current technology. Summary of the Invention

[0010] The purpose of this invention is to at least partially solve some of the technical problems existing in the prior art, and to provide an autonomous operation control method and system for a mining intelligent depressurization machine. Its structure is reasonable, and it constructs a closed loop of autonomous operation throughout the entire process of "perception-decision-execution-evaluation", so as to realize the unmanned and intelligent operation of depressurization.

[0011] To solve the above-mentioned technical problems, the present invention provides an autonomous operation control method for a mine intelligent depressurization machine, comprising the following steps: Real-time image data of the surrounding rock in the tunnel is acquired, and the image data is inferred based on the YOLOv11 model to identify high-risk areas and their corresponding spatial location information. By integrating lidar point cloud data and visual image data, a three-dimensional semantic map of the alleyway is constructed, and an autonomous walking path is planned based on the spatial location information of the high-risk area. The pressure relief machine is controlled to move along the autonomous walking path to the target work point, and the drill rod rotation mechanism and telescopic mechanism are controlled to perform drilling operations according to the preset pressure relief parameters. When drilling reaches the preset depth, the hole reaming command is automatically triggered, controlling the umbrella-type hydraulic hole reaming mechanism to open radially and perform variable diameter hole reaming operation; Acquire ultrasonic imaging data and surrounding rock microseismic monitoring data inside the borehole, invert the stress distribution around the borehole, and generate an evaluation result of the pressure relief effect.

[0012] In some embodiments, the training dataset of the YOLOv11 model includes roadways with different cross-sectional shapes, different support methods, and on-site images before and after rockbursts. The model output includes the identification results of support damage areas, rock mass detachment areas, and structural instability areas.

[0013] In some embodiments, constructing a three-dimensional semantic map of a tunnel specifically includes: generating a three-dimensional point cloud map using a laser SLAM algorithm, extracting semantic features using a visual SLAM algorithm, and fusing semantic labels with the three-dimensional point cloud to form a semantic map containing support structures and hazardous area markers.

[0014] In some embodiments, the autonomous walking path is planned using the Dijkstra algorithm, which generates a globally optimal path with the current location of the equipment as the starting point, the target work point as the ending point, and the alleyway passage area as the constraint condition. During the walking process, a local obstacle avoidance algorithm is used to avoid sudden obstacles.

[0015] In some embodiments, the variable diameter reaming operation includes: maintaining a conventional borehole diameter in the shallow surrounding rock section, and controlling the umbrella-type hydraulic reaming mechanism to open step by step according to the preset diameter value after drilling to a preset depth, thereby achieving deep reaming. The preset diameter value is related to the roadway support parameters and the stress inversion results of the surrounding rock.

[0016] In some embodiments, the timing of the hole enlargement command is dynamically adjusted based on the drilling depth, drill rod torque change and surrounding rock hardness. When the drilling depth reaches a set threshold and the torque change meets preset conditions, the hole enlargement action is automatically executed.

[0017] In some embodiments, inverting the stress distribution around the borehole includes: identifying the range of the fractured zone and the plastic zone inside the borehole based on ultrasonic imaging data, combining the waveform characteristics and energy release parameters in the microseismic monitoring data, calculating the stress field distribution around the borehole through a stress inversion model, and comparing it with the theoretical stress distribution model.

[0018] In some embodiments, the autonomous operation control method further includes a safety interlock control step: before drilling operations, the locking status of the hydraulic support system, the positioning status of the drill rod, and the ambient gas concentration are automatically checked, and drilling operations are prohibited if any condition is not met; during the operation, when the concentration of explosive gas is detected to exceed the limit, the power supply to the non-intrinsically safe circuit is immediately cut off and the equipment is controlled to automatically retreat to a safe area.

[0019] Furthermore, this invention also discloses an autonomous operation control system for a mining intelligent depressurization machine, comprising: The sensing module includes a high-definition camera, lidar, ultrasonic detector and micro-vibration sensor, used to collect tunnel images, point cloud data, borehole imaging and surrounding rock vibration data; The main control decision module, equipped with the YOLOv11 model and ROS2 architecture, is used to identify high-risk areas, build semantic maps, plan autonomous walking paths, and generate drilling and reaming control commands. The execution control module includes drivers for the drill rod rotation mechanism, the telescopic mechanism, and the umbrella-type hydraulic reaming mechanism, used to receive and execute the control commands; The evaluation feedback module is used to invert the stress distribution around the hole and generate an evaluation result of the pressure relief effect, which is then fed back to the main control decision module to optimize subsequent operation parameters.

[0020] In some embodiments, the master control decision module includes an intrinsically safe industrial computer for mining and a lower-level real-time controller. The industrial computer runs the ROS2 architecture and undertakes visual recognition and path planning algorithm calculations. The lower-level real-time controller communicates with the execution control module through the CANopen bus to realize real-time closed-loop control of drill rod feeding, hole reaming action and walking drive.

[0021] Compared with the prior art, the present invention has the following beneficial effects: a. Constructing a closed-loop autonomous operation process encompassing "perception—decision-execution—evaluation" to achieve unmanned and intelligent pressure relief operations. This invention integrates the YOLOv11 visual recognition model with the ROS2 architecture to achieve fully autonomous control from high-risk area identification to pressure relief effect evaluation. Specifically, high-definition cameras acquire tunnel images in real time; the YOLOv11 model automatically identifies high-risk areas such as support damage and rockfall and outputs spatial location information; laser SLAM and visual SLAM are fused to construct a 3D map with semantic labels, and the Dijkstra algorithm is used to plan the globally optimal path, allowing the equipment to autonomously travel to the target work point; during drilling, the drill rod rotation and extension are automatically controlled according to preset pressure relief parameters, and a hole-reaming command is automatically triggered at a preset depth, controlling the umbrella-type hydraulic hole-reaming mechanism to perform variable-diameter hole-reaming operations; after the operation is completed, the stress distribution around the hole is inverted using ultrasonic imaging and microseismic monitoring data to generate a pressure relief effect evaluation result. The aforementioned technologies together form a complete autonomous operation closed loop, requiring no human intervention throughout the entire process. This fundamentally eliminates the safety risks for operators working in high-stress areas and achieves an intelligent upgrade of "machine replacing human."

[0022] b. Establish a data-driven dynamic optimization mechanism for variable diameter operation parameters to improve the accuracy and adaptability of pressure relief. This invention breaks through the limitations of traditional variable diameter pressure relief operations that rely on empirical parameters, and establishes a dynamic optimization mechanism for parameters based on real-time monitoring data. On the one hand, the triggering timing of the hole enlargement command no longer depends solely on the preset depth, but is dynamically judged by comprehensively considering multi-source information such as drilling depth, drill rod torque changes, and surrounding rock hardness. When the drilling depth reaches the set threshold and the torque change meets the preset conditions, the hole enlargement action is automatically executed, matching the hole enlargement timing with the actual state of the surrounding rock. On the other hand, the preset diameter value is correlated with the roadway support parameters and the results of surrounding rock stress inversion. The pressure relief effect evaluation results are fed back to the main control decision module through the evaluation feedback module to optimize parameters such as the hole enlargement diameter at subsequent operation points. When the spacing of the anchor bolt rows changes in different sections of the same roadway, the system can automatically adjust the hole enlargement diameter to adapt to the differentiated pressure relief needs without manual intervention or replacement of mechanical parts. The above mechanism transforms the pressure relief parameters from "static preset" to "dynamic optimization," significantly improving the accuracy of variable diameter pressure relief and its adaptability to different geological conditions.

[0023] c. Constructing a multi-level safety interlocking and explosion-proof control system to ensure operational reliability under harsh underground conditions. This invention addresses the high dust, high humidity, strong vibration, and explosive gas environments underground by constructing a multi-level safety interlocking system with both hardware and software collaboration. Before operation, the system automatically verifies the locking status of the hydraulic support system, the drill pipe positioning status, and the ambient gas concentration. Drilling operations are prohibited if any condition is not met, mitigating operational risks from the outset. During operation, the explosive gas concentration is monitored in real time. If the concentration exceeds the limit, the power supply to non-intrinsically safe circuits is immediately cut off, and the equipment is automatically withdrawn to a safe area, achieving proactive risk avoidance under environmental risks. Simultaneously, this invention adopts a dual-core redundant architecture of a mining intrinsically safe industrial computer and a lower-level real-time controller. Control and safety functions are physically isolated and operate independently; a single module failure does not affect the system's safety protection capabilities. The electrical system strictly adheres to explosion-proof and intrinsically safe design requirements, with electrical isolation between high-voltage and low-voltage circuits to ensure safe operation in explosive environments. These designs enable this invention to achieve stable and reliable autonomous operation under harsh underground conditions, providing key equipment support for safe production in intelligent mines. Attached Figure Description

[0024] The advantages of the present invention will become clearer and more readily understood through the following detailed description in conjunction with the accompanying drawings, which are merely illustrative and do not limit the invention, wherein: Figure 1 This is a schematic diagram of the autonomous operation control method and system of a mining intelligent pressure relief machine according to the present invention; Figure 2 This is a flowchart of an autonomous operation control method for a mining intelligent depressurization machine according to the present invention; Figure 3 This is a schematic diagram of the sensing module described in this invention; Figure 4 This is a schematic diagram of the operation principle of the execution control module described in this invention. Detailed Implementation

[0025] Figures 1 to 4 This is a schematic diagram of the autonomous operation control method and system of a mining intelligent depressurization machine described in this application. The invention will be described in detail below with reference to specific embodiments and accompanying drawings.

[0026] The embodiments described herein are specific implementations of the present invention, used to illustrate the concept of the invention, and are illustrative and exemplary, and should not be construed as limiting the implementation or scope of the invention. In addition to the embodiments described herein, those skilled in the art can employ other obvious technical solutions based on the content disclosed in the claims and specification of this application. These technical solutions include those that make any obvious substitutions and modifications to the embodiments described herein.

[0027] The accompanying drawings in this specification are schematic diagrams used to illustrate the concept of the invention, and schematically show the shapes of the various parts and their interrelationships. Please note that, in order to clearly show the structure of the components in the embodiments of the invention, the drawings are not drawn to the same scale. The same reference numerals are used to indicate the same parts.

[0028] Example 1: This example provides an autonomous operation control method and system for a mine intelligent pressure relief machine. For example... Figure 1 As shown, the system includes a depressurization machine body 1000, a sensing module 1500, a main control decision module 2000, an execution control module 3000, and an evaluation feedback module 4000.

[0029] The pressure relief machine body 1000 adopts a tracked walking structure, and its bottom is equipped with... Figure 4 The tracked traveling mechanism 1010 shown includes two independently driven tracks, one on the left and one on the right. Each track is equipped with an independent servo drive motor, enabling forward, backward, and on-the-spot turning. Hydraulic support systems 1020 are located at the four corners of the vehicle body, including a front hydraulic support 1021 and a rear hydraulic support 1022. These systems provide ground support during drilling operations to counteract the drill bit's counter-torque and can raise the vehicle body to accommodate drilling operations at different heights. The drill rod system includes a drill rod rotation mechanism 1100, a telescopic mechanism 1200, a drilling mechanism 1300, and a umbrella-type hydraulic reaming mechanism 1400.

[0030] The drill pipe rotation mechanism 1100 is mounted at the front of the vehicle body and includes a rotary drive motor 1110 and a rotary angle encoder 1120. It is used to rotate the drill pipe from its horizontal storage position during inspection to its vertical position in the surrounding rock during operation, with a rotation angle control accuracy of ±0.5°. The telescopic mechanism 1200 adopts a multi-stage hydraulic telescopic rod 1210 series structure, with a total length of 2.2m in the retracted state and up to 8m when extended. A wire-type displacement sensor 1220 is used to collect the telescopic length in real time, with a control accuracy of ±2cm. The drilling mechanism 1300 includes a PDC drill bit 1310 and a drilling drive motor 1320. The drilling drive motor 1320 has a built-in torque sensor 1330 and a speed sensor, realizing dual closed-loop control of torque and speed. It can adjust the drill bit speed and drilling torque in real time according to the hardness of the surrounding rock. The umbrella-type hydraulic reaming mechanism 1400 is mounted at the front end of the drill pipe; for its specific structure, please refer to the aforementioned utility model patent embodiment.

[0031] The sensing module 1500 includes a high-definition camera 1510, a lidar 1520, an ultrasonic detector 1530, a micro-vibration sensor 1540, and a gas sensor array 1550. The high-definition camera 1510 is mounted on the top front of the vehicle body, employing a mining-grade intrinsically safe high-definition wide-angle camera covering a 150° field of view, and is equipped with a high-power supplementary lighting system for acquiring real-time image data of the surrounding rock in the tunnel. The lidar 1520 is a mining-grade intrinsically safe multi-line lidar used to acquire three-dimensional point cloud data of the tunnel. The ultrasonic detector 1530 is mounted on the front end of the telescopic mechanism 1200, used for ultrasonic scanning imaging of the borehole interior during drilling, acquiring data on borehole porosity, surrounding rock displacement, and the extent of fractured zones. The micro-vibration sensor 1540 wirelessly communicates with pre-embedded micro-vibration monitoring nodes within the tunnel, acquiring surrounding rock vibration waveform data. The gas sensor array 1550 includes sensors for methane, carbon monoxide, carbon dioxide, and oxygen concentrations, as well as temperature, humidity, and dust sensors, used for real-time monitoring of underground environmental parameters.

[0032] The main control decision module 2000 adopts a dual-core redundant architecture of upper-level computer + lower-level computer. The upper-level computer is a mining-grade intrinsically safe industrial computer 2100, equipped with a multi-core high-performance processor and a large-capacity industrial-grade solid-state drive, running the ROS2 architecture. It is responsible for intelligent algorithm calculations such as YOLOv11 model inference, SLAM mapping, and path planning, and stores the tunnel model, training dataset, operation logs, and monitoring data. The lower-level computer includes a Raspberry Pi 4 Model B as the main controller and a mining-grade intrinsically safe PLC 2300 as a backup controller for safety interlocks. It communicates with the execution control module 3000 through the CANopen bus to realize high-precision servo control and safety interlock logic of the actuators.

[0033] The execution control module 3000 includes a driver for the drilling drive motor 1320, a driver for the rotary drive motor 1110, a servo driver for the tracked walking mechanism 1010, an electro-hydraulic proportional control valve group for the hydraulic support system 1020, an electro-hydraulic proportional control valve group for the telescopic mechanism 1200, and a hydraulic servo valve group for the umbrella-type hydraulic reaming mechanism 1400.

[0034] The evaluation feedback module 4000 is integrated into the industrial computer 2100. It is used to invert the stress distribution around the hole based on ultrasonic imaging data and microseismic monitoring data, generate the pressure relief effect evaluation result, and feed it back to the main control decision module 2000 to optimize subsequent operation parameters.

[0035] The ground monitoring terminal 5000 establishes two-way communication with the depressurization machine through a mining intrinsically safe 5G / WiFi6 wireless communication gateway, enabling real-time uploading of operation data, monitoring videos, and equipment status, as well as remote command issuance.

[0036] Example 2: Autonomous Operation Control Method Flowchart. This example details the specific flowchart of the autonomous operation control method for the intelligent pressure relief machine used in mining. The detailed flowchart is shown below. Figure 2 As shown.

[0037] Step S101: High-risk area identification During the autonomous inspection process of the depressurization machine within the tunnel, a high-definition camera 1510 acquires real-time image data of the surrounding rock at a rate of 30 frames per second. The image resolution is 1920×1080 pixels, and the data is transmitted to an industrial computer 2100 via gigabit Ethernet. The industrial computer 2100 loads a pre-trained YOLOv11 model to perform inference on the image data.

[0038] The training dataset for the YOLOv11 model was collected and constructed by our team, covering roadways with different cross-sectional shapes such as rectangular, arched, and horseshoe-shaped sections; different support methods such as bolt support, mesh support, shotcrete support, and scaffold support; and on-site images of typical working conditions before and after rockbursts, support system damage, and rock mass detachment, totaling 12,600 images. The dataset was manually labeled, with labeled categories including: support damage areas (bolt breakage, tray detachment, mesh tearing), rock mass detachment areas (slab spalling, rockfall), and structural instability areas (roof subsidence, side bulging). The model was trained using data augmentation techniques (including random flipping, rotation, scaling, and brightness adjustment). After 10,000 iterations, the model achieved an accuracy of 86% and a recall of 82%.

[0039] When the YOLOv11 model detects a high-risk area, it outputs the recognition result and the corresponding spatial location information. The spatial location information is converted using the camera pose provided by the visual SLAM algorithm to obtain the coordinates of the high-risk area in the three-dimensional coordinate system of the tunnel. The recognition result is uploaded in real time to the ground monitoring terminal 5000 for early warning display via wireless communication.

[0040] Step S102: 3D semantic map construction and path planning The industrial computer 2100 simultaneously receives 3D point cloud data (300,000 points per second) collected by the lidar 1520 and image data collected by the high-definition camera 1510, and performs fusion mapping using the SLAM algorithm under the ROS2 architecture.

[0041] Specifically, the laser SLAM algorithm employs an improved LOAM (Lidar Odometry and Mapping) algorithm to extract features (including corner features and planar features) from point cloud data. It then calculates the lidar odometry through inter-frame registration to construct a 3D point cloud map. The visual SLAM algorithm uses the ORB-SLAM3 algorithm to extract ORB feature points from the image. It constructs a sparse feature map through feature matching and triangulation, and utilizes a deep learning model to perform semantic segmentation on the image, extracting semantic labels such as support structures, tunnel boundaries, and hazardous area markers.

[0042] Semantic labels are integrated with 3D point cloud maps: Semantic labels are projected onto the 3D point cloud space through extrinsic parameter calibration of cameras and LiDAR, and each point in the point cloud is assigned semantic attributes (such as "anchor bolt", "mesh", "rock mass"), forming a 3D semantic map containing semantic information. In the semantic map, passageways are marked as passable areas, and high-risk areas are marked as prohibited areas or areas of special concern.

[0043] Based on the spatial location information of high-risk areas, the industrial computer 2100 uses Dijkstra's algorithm to plan its autonomous movement path. Specifically, taking the current location of the equipment as the starting point and a predetermined work point near the high-risk area as the ending point, and using the passageway area as a constraint, a weighted directed graph is constructed (nodes are discrete locations within the passageway, edges are passable paths, and weights are path lengths). The shortest path from the starting point to the ending point is calculated as the globally optimal path. The global path planning result is sent to the lower-level real-time controller 2200 via the CANopen bus.

[0044] As the equipment moves along the global path, the lidar 1520 scans the environment ahead in real time, and the lower-level real-time controller 2200 runs a local obstacle avoidance algorithm (dynamic window method) to avoid sudden obstacles (such as falling rocks or parked equipment) in real time, generate a local trajectory, and control the tracked walking mechanism 1010 to execute it.

[0045] Step S103: Autonomous Drilling Operation After the pressure relief machine moves to the target work point along the autonomous walking path, it performs a safety interlock check before drilling. The lower-level real-time controller 2200 automatically checks the following conditions: (1) Locking status of the hydraulic support system 1020: both the front hydraulic support 1021 and the rear hydraulic support 1022 have been deployed and are in contact with the ground, and the pressure sensor feedback shows that the support force meets the preset threshold; (2) Drill rod positioning status: the rotary angle encoder 1120 feedback shows that the drill rod has been rotated to the vertical position of the surrounding rock, and the mechanical locking mechanism is in place; (3) Ambient gas concentration: the methane concentration detected by the gas sensor group 1550 is less than 0.5%, and the oxygen concentration is in the range of 19.5% to 23.5%. If any condition is not met, drilling is prohibited from starting and an audible and visual alarm is issued.

[0046] After successful verification, the industrial computer 2100 generates drilling control commands based on preset pressure relief parameters. These preset parameters include: conventional drilling depth H1 (e.g., 5m) and conventional drilling diameter D1 (e.g., 75mm) in the shallow surrounding rock section; depth H2 (e.g., 10m) and reaming diameter D2 (e.g., 150mm) in the deep reaming section. The control commands are sent to the lower-level real-time controller 2200 via the CANopen bus, which further controls the execution control module 3000 to execute them.

[0047] The drilling operation process is as follows: First, the multi-stage hydraulic telescopic rod 1210 of the telescopic mechanism 1200 extends, and the wire-type displacement sensor 1220 provides real-time feedback on the extension length, controlling the drill rod feed to the predetermined starting position. Simultaneously, the drilling drive motor 1320 starts, driving the PDC drill bit 1310 to rotate. The torque sensor 1330 and speed sensor monitor the drilling torque and speed in real time, and the lower-level real-time controller 2200 adaptively adjusts the drilling parameters according to the surrounding rock hardness to achieve constant power drilling. In the shallow surrounding rock section (0~5m), the umbrella-type hydraulic reaming mechanism 1400 remains in a retracted state, and the drill bit drills with a conventional diameter of 75mm.

[0048] When the drilling depth reaches 5m, the drilling depth signal fed back by the encoder triggers the hole reaming preparation command. The lower-level real-time controller 2200 comprehensively judges the drilling depth (≥5m), the change in drill rod torque (torque decreases from the peak and then tends to stabilize), and the hardness of the surrounding rock (estimated based on torque and feed rate). When the torque change meets the preset conditions, the hole reaming command is automatically triggered.

[0049] Step S104: Variable Diameter Reaming Operation The reaming command is issued to the hydraulic servo valve assembly of the umbrella-type hydraulic reaming mechanism 1400. The hydraulic servo valve assembly controls the hydraulic cylinder 1431 to drive the connecting rod slider 1432 to move forward. The groove structure on the surface of the connecting rod slider 1432 causes the reaming vane 1420 to open radially. The displacement sensor 1440 detects the displacement of the connecting rod slider 1432 in real time, and the angle sensor 1450 detects the opening angle of the reaming vane 1420 in real time. The hydraulic servo valve assembly performs PID closed-loop control according to the preset target value of the reaming diameter (150mm) to ensure that the reaming vane 1420 is accurately opened to the corresponding angle.

[0050] The PDC composite cutting teeth 1421 on the outer edge of the reaming blade 1420 contact the borehole wall and cut the surrounding rock under the rotation of the drill pipe, increasing the borehole diameter from 75mm to 150mm. During the reaming process, the torque feedback of the drilling drive motor 1320 is used to determine the reaming cutting status. When the torque increases abnormally, the feed rate is automatically reduced to prevent the drill from getting stuck.

[0051] After the deep reaming section (5m~10m) is completed, the hydraulic servo valve group controls the hydraulic cylinder 1431 to reverse, the connecting rod slider 1432 moves backward, and the reaming vane 1420 retracts to the contracted state. Then the telescopic mechanism 1200 slowly retracts, and the drill bit exits the borehole. The rotary drive motor 1110 rotates the drill rod back to the horizontal storage position.

[0052] Step S105: Evaluation of pressure relief effect During and after drilling and reaming operations, the ultrasonic detector 1530 emits ultrasonic waves in real time to scan and image the interior of the borehole, acquiring data such as the porosity distribution, surrounding rock displacement, and fracture zone extent on the borehole inner wall. The ultrasonic detector 1530 employs phased array ultrasonic technology, enabling 360° scanning of the entire borehole circumference with an imaging resolution of 0.5 mm. The imaging data is transmitted to the industrial computer 2100 via gigabit Ethernet.

[0053] Simultaneously, the microseismic sensor 1540 wirelessly communicates with pre-embedded microseismic monitoring nodes within the tunnel to collect surrounding rock vibration waveform data, including event occurrence time, waveform amplitude, frequency characteristics, and energy release parameters. The microseismic data is sampled at a frequency of 1000Hz and uploaded in real-time via a 4G / 5G wireless network.

[0054] The evaluation feedback module 4000 in the industrial computer 2100 performs the following operations: First, the ranges of the fractured and plastic zones within the borehole are identified based on ultrasonic imaging data. The fractured areas on the borehole wall are extracted using an image segmentation algorithm, and the radii of the fractured zone (R_f) and the plastic zone (R_p) are calculated. The observed radii of the fractured zone are compared with the theoretically calculated values ​​to determine whether the pressure relief is sufficient.

[0055] Secondly, combining waveform characteristics and energy release parameters from microseismic monitoring data, the stress field distribution around the borehole is calculated using a stress inversion model. The stress inversion model is based on moment tensor theory, utilizing the initial polarity and amplitude of P-waves and S-waves recorded by multiple microseismic sensors to invert the source mechanism and then calculate the surrounding rock stress tensor. The inversion results are compared with the aforementioned theoretical stress distribution model (a multi-borehole mechanical model based on an improved Kirsch solution) to evaluate the actual effect of stress transfer after decompression.

[0056] Finally, a pressure relief effect evaluation report is generated, including indicators such as the location and magnitude of the peak stress around the borehole after pressure relief, the stress transfer distance, and the degree of impact on the support system. The evaluation results are uploaded wirelessly to the ground monitoring terminal 5000 and fed back to the main control decision module of the industrial computer 2100 for optimizing the pressure relief parameters at subsequent work sites.

[0057] Step S106: Safety Interlock and Emergency Handling Throughout the entire operation, the safety interlock and protection mechanisms operate continuously.

[0058] Safety interlock before operation: As mentioned above, the system automatically checks the locking status of the hydraulic support system 1020, the positioning status of the drill rod, and the ambient gas concentration. Drilling operations are prohibited if any condition is not met.

[0059] Safety interlocking during operation: The lower-level real-time controller 2200 and the intrinsically safe mining PLC 2300 monitor equipment status and environmental parameters in parallel. When the gas sensor group 1550 detects that the concentration of explosive gas exceeds the limit (methane concentration ≥1.0%), the intrinsically safe mining PLC 2300 immediately cuts off the power supply to the non-intrinsically safe circuit, stops all operations, switches to the intrinsically safe backup power supply, triggers an audible and visual alarm, and controls the equipment to automatically retreat to a safe area. The retreat path is planned in real time by the industrial computer 2100 based on the current equipment location and the roadway map, and executed by the lower-level real-time controller 2200.

[0060] Mechanism action interlock: When the drill rod is not retracted into position, the hole reaming mechanism is not retracted, or the hydraulic support is not retracted, the system prohibits the equipment from traveling at high speed to prevent equipment damage caused by conflicting mechanism actions.

[0061] Emergency Stop Interlock: The equipment body is equipped with multiple intrinsically safe emergency stop buttons, and the ground monitoring terminal 5000 is equipped with a remote emergency stop function. When any emergency stop is triggered, the intrinsically safe mining PLC2300 immediately cuts off the power supply to the high-power circuit, locks all actuators, stops all operations, and triggers an audible and visual alarm.

[0062] Example 3: Dynamic optimization of hole enlargement triggering timing This embodiment, based on embodiment 2, provides a detailed explanation of the dynamic optimization of the timing for triggering the hole enlargement command.

[0063] In existing technologies, the timing of borehole reaming triggering typically relies solely on a preset depth, which cannot adapt to changes in the hardness of the surrounding rock. This embodiment achieves dynamic optimization of borehole reaming timing by integrating multi-source information.

[0064] Specifically, during the drilling process, the lower-level real-time controller 2200 records multi-dimensional data such as drilling depth d, drilling torque T, feed speed v, and drill rod vibration frequency f in real time. When the drilling depth reaches the preset threshold d... th When the borehole reaches a depth of 5m (for example), the system enters the borehole reaming preparation state. At this time, the system begins to calculate the torque change rate ΔT / Δt and the surrounding rock hardness index H. rock .

[0065] Rock hardness index H rock The drilling torque T and the uniaxial compressive strength σ of the surrounding rock are estimated as follows: Under standard drilling conditions (constant feed rate v), the drilling torque T and the uniaxial compressive strength σ of the surrounding rock are estimated. c There is a positive correlation. The system collects the average torque T during the most recent 1m drilling process. avg , compared with the pre-calibrated standard surrounding rock torque T ref Compare and calculate the hardness index H. rock = T avg / T ref When H rock When H > 1.2, it is classified as hard rock. rock A value less than 0.8 indicates soft rock.

[0066] The comprehensive judgment of the hole enlargement trigger condition is as follows: When d ≥ d th The system will automatically trigger a hole enlargement command when any of the following conditions are met: (1) The torque change rate ΔT / Δt changes from positive to negative and the absolute value is greater than the preset threshold, indicating that the drill bit has passed through the high stress zone and entered the relatively relaxed region; (2) Torque fluctuation variance σ_T 2 When the temperature drops below the preset threshold, it indicates that the surrounding rock is becoming more uniform and is suitable for borehole enlargement operations. (3) The feed rate v at the drilling depth d th The step increase in the surrounding area indicates a significant decrease in the strength of the surrounding rock.

[0067] For hard rock surroundings, the system automatically postpones the borehole enlargement trigger depth to d. th + Δd (Δd=0.5~1.0m) to avoid premature hole enlargement in hard surrounding rock, which could lead to excessive wear of the cutting teeth. For soft surrounding rock, the system automatically advances the hole enlargement trigger depth to d_th - Δd to ensure timely stress transfer in shallow surrounding rock sections.

[0068] Through the above dynamic optimization, this embodiment enables the timing of borehole enlargement to be precisely matched with the actual state of the surrounding rock, effectively avoiding problems such as poor pressure relief or equipment damage caused by borehole enlargement too early or too late.

[0069] Example 4: Closed-loop optimization of borehole diameter This embodiment, based on Embodiment 2, provides a detailed explanation of the closed-loop optimization of the hole diameter.

[0070] In existing technologies, the diameter of the enlarged borehole is usually statically preset based on roadway support parameters (such as the spacing between anchor bolt rows), and cannot be dynamically adjusted according to the actual pressure relief effect. In this embodiment, the pressure relief effect evaluation result is fed back to the main control decision module 2000 through the evaluation feedback module 4000, thereby realizing closed-loop optimization of the enlarged borehole diameter.

[0071] Specifically, for the first work point in a certain section of the roadway, the system uses a preset hole enlargement diameter D. preset (For example, 150mm) Perform a variable diameter reaming operation. After the operation is completed, the evaluation feedback module 4000 inverts the stress distribution around the hole based on ultrasonic imaging data and microseismic monitoring data, and calculates the following indicators: (1) Stress transfer distance L transfer The distance between the peak stress location around the hole and the boundary of the original rock stress zone; (2) Stress relief rate η release : (Original rock stress σ) original - Residual stress σ around the hole residual ) / σ original ; (3) Radius of the crushing zone R f With the radius R of the plastic zone p The ratio of .

[0072] Compare the above indicators with the theoretical target value: If the stress transfer distance is insufficient (L) transfer <L target ) or low stress release rate (η) release < η target This indicates insufficient pressure relief. The evaluation feedback module 4000 generates optimization suggestions, increasing the borehole diameter at the next work point by ΔD (e.g., by 10mm). If the ratio of the crushing zone radius to the plastic zone radius is too large (R... f / R p A value >0.8 indicates that there may be excessive pressure relief, and the system will reduce the orifice diameter by ΔD at the next working point.

[0073] The optimized borehole diameter parameters are automatically updated in the operation parameter library of the main control decision module 2000, and subsequent operation points are executed according to the optimized parameters. When the spacing between anchor bolt rows changes in different sections of the same roadway, the system automatically adjusts the preset diameter benchmark value according to the input support parameters of different sections, and performs the above closed-loop optimization at the first operation point of each section, so that the borehole diameter quickly converges to the optimal value.

[0074] Through the aforementioned closed-loop optimization mechanism, this embodiment transforms the borehole diameter from "static preset" to "dynamic optimization," significantly improving the accuracy of variable diameter pressure relief and its adaptability to different support conditions.

[0075] Example 5: Stress Inversion Method Based on Multi-Source Information Fusion This embodiment provides a detailed description of the stress inversion method in step S105.

[0076] The stress inversion method is based on the moment tensor theory and combines ultrasonic imaging data with microseismic monitoring data for joint inversion.

[0077] First, microseismic events in the surrounding rock were acquired using a microseismic sensor 1540. For each microseismic event, the arrival times t of the P-wave and S-wave at multiple sensors were recorded. Pi t Si The system obtains information such as the initial polarity and amplitude of the P-wave. It locates the seismic source (x0, y0, z0) and the time of origin t0 by solving the following set of equations: t_Pi - t0 = √((x_i - x0)^2 + (y_i - y0)^2 + (z_i - z0)^2) / v_P t_Si - t0 = √((x_i - x0)^2 + (y_i - y0)^2 + (z_i - z0)^2) / v_S Where v_P is the P-wave velocity and v_S is the S-wave velocity, which are preset according to the lithology of the surrounding rock.

[0078] After the seismic source is located, the focal mechanism (tensional, shear, or hybrid) is determined based on the initial polarity of the P-wave, and the moment tensor M is calculated. The moment tensor M is a 3×3 symmetric tensor, and its components are related to the stress field as follows: M = μA u Where μ is the shear modulus, A is the fracture area, and u is the average sliding displacement.

[0079] The principal stress directions and stress release amounts can be calculated using the moment tensor M. Combining the fractured zone R_f and plastic zone R_p identified by ultrasonic imaging data, the stress field distribution around the borehole is inverted using the stress-strain relationship. Specifically, according to elasticity theory, the stress field around the borehole can be expressed as: σ_r = (σ_1 + σ_3) / 2 * (1 - R_f^2 / r^2) + (σ_1 - σ_3) / 2 * (1 - 4R_f^2 / r^2 + 3R_f^4 / r^4) * cos2θ σ_θ = (σ_1 + σ_3) / 2 * (1 + R_f^2 / r^2) - (σ_1 - σ_3) / 2 * (1 + 3R_f^4 / r^4) * cos2θ Where σ_1 and σ_3 are the maximum and minimum principal stresses, r is the distance from the borehole center, and θ is the polar angle.

[0080] By using the principal stress directions and magnitudes obtained from the moment tensor inversion as input boundary conditions and substituting them into the above formula, the stress distribution at any location around the hole can be calculated. The calculation results are compared with the theoretical stress distribution model (a multi-hole mechanical model based on the Kirsch solution) to generate a stress inversion report.

[0081] Through the stress inversion method of multi-source information fusion described above, this embodiment realizes a quantitative evaluation of the pressure relief effect, providing a scientific basis for subsequent optimization of operating parameters.

[0082] Example 6: Explosion-proof and Intrinsically Safe Design This embodiment provides a detailed description of the system's explosion-proof and intrinsically safe design.

[0083] The system strictly adheres to the design specifications for electrical equipment in explosive atmospheres in underground coal mines (GB 3836 series standards), and its core explosion-proof and intrinsically safe designs are as follows: Explosion-proof design: High-power electrical components, including the drilling drive motor 1320, rotary drive motor 1110, servo motor of the tracked walking mechanism 1010, generator, main power distribution equipment, etc., adopt a mining-grade explosion-proof design (Exd I Mb). The explosion-proof enclosure is made of high-strength cast steel or welded steel plate structure, with a wall thickness of not less than 10mm, a gap between explosion-proof mating surfaces of not more than 0.2mm, and a mating surface length of not less than 25mm. The explosion-proof enclosure can withstand an internal explosion pressure of 1.5MPa without permanent deformation or rupture, and extinguishes the flame through the explosion-proof mating surfaces, preventing the explosion from propagating to the outside.

[0084] Intrinsically safe design: The low-voltage control components include a main control decision module 2000, a sensing module 1500 (high-definition camera 1510, lidar 1520, ultrasonic detector 1530, micro-vibration sensor 1540, and gas sensor group 1550, such as...) Figure 3As shown), the communication module, etc., adopt the intrinsically safe (Exia 1 Mb) design for mining applications. The intrinsically safe circuit strictly limits the voltage to ≤24V, current to ≤100mA, capacitance to ≤1μF, and inductance to ≤1mH, ensuring that under normal operation and fault conditions, neither electrical sparks nor thermal effects can ignite the explosive gas mixture.

[0085] Electrical isolation: Intrinsically safe circuits and non-intrinsically safe circuits are reliably electrically isolated and energy-limited using intrinsically safe safety barriers, with an isolation voltage ≥1500V, to prevent dangerous energy from non-intrinsically safe circuits from entering intrinsically safe circuits. Intrinsically safe and non-intrinsically safe circuits are physically separated by a minimum spacing of 50mm to avoid cross-wiring and electromagnetic coupling.

[0086] Electromagnetic Compatibility: The entire circuit uses shielded cables with reliable double-end grounding of the shielding layer. The power supply circuit is equipped with a filter module to suppress conducted interference. The signal circuit is equipped with surge suppression and filtering circuits to prevent interference from entering. A single-point grounding design is adopted to reduce ground loop interference.

[0087] Environmental protection: All electrical components have an IP65 protection rating, suitable for underground working conditions with high dust, high humidity, and strong vibration. The electrical equipment installation adopts a vibration-damping design, with a natural frequency of ≤10Hz for the dampers. Precision components are equipped with impact-resistant and dustproof protective housings, made of 3mm thick welded steel plates with a rust-proof paint coating.

[0088] Through the aforementioned explosion-proof and intrinsically safe design, this embodiment ensures the safe and reliable operation of the system in working environments with high gas and dust explosion risks, such as underground coal mines.

[0089] Example 7: Overall System Workflow The following section provides a comprehensive description of the implementation methods of this invention, using a complete workflow example.

[0090] A coal mine roadway has a rectangular cross-section, 5m wide and 3.5m high, supported by anchor bolts with a row spacing of 0.8m. The roadway has been excavated and requires drilling for pressure relief. The ground monitoring terminal 5000 issues a work order to the pressure relief machine, including: roadway starting coordinates (0,0,0), ending coordinates (100,0,0), work point spacing of 1.6m (i.e., one pressure relief hole between every two rows of anchor bolts), shallow drilling depth of 5m, shallow drilling diameter of 75mm, deep reaming depth of 10m, and deep reaming diameter of 150mm.

[0091] The decompression machine autonomously moves along the tunnel to its starting position. A high-definition camera (1510) captures images of the tunnel, and the YOLOv11 model detects a support damage area (anchor bolt tray detached) 10m ahead, outputting the coordinates of the high-risk area (10, 1.5, 1.2). The main control decision module (2000) marks the high-risk area as a priority area and replans the path, avoiding the area within 2m below the high-risk zone.

[0092] The machine reaches the first working point (2, 1.5, 1.2). The hydraulic support system 1020 deploys, with the front hydraulic support 1021 and the rear hydraulic support 1022 supporting the ground. The pressure sensor reports a support force of 50kN. The rotary drive motor 1110 drives the drill rod to rotate to a position perpendicular to the surrounding rock. The rotation angle encoder 1120 reports an angle of 90°±0.2°. The gas sensor group 1550 detects a methane concentration of 0.2% and an oxygen concentration of 20.5%, which meet the requirements.

[0093] Drilling operation begins. The multi-stage hydraulic telescopic rod 1210 extends, the wire-type displacement sensor 1220 provides feedback on the length, and the drill rod is fed to the borehole position. The drilling drive motor 1320 starts, and the PDC drill bit 1310 drills at a speed of 1500 rpm and a feed rate of 50 mm / min. The torque sensor 1330 monitors the drilling torque in real time, maintaining the drilling torque at 200-250 N·m in the shallow surrounding rock section of 0-5m.

[0094] When the drilling depth reaches 5m, the system enters the reaming preparation state. The average drilling torque over the most recent 1m is calculated to be 220 N·m, indicating that the surrounding rock hardness is moderate. When the drilling depth reaches 5.2m, the torque change rate changes from positive to negative, and the system automatically triggers the reaming command.

[0095] The umbrella-type hydraulic reaming mechanism 1400 is activated. The hydraulic servo valve group controls the hydraulic cylinder 1431 to drive the connecting rod slider 1432 to move forward. The displacement sensor 1440 provides feedback on the displacement, and the angle sensor 1450 provides feedback on the opening angle of the reaming vane 1420. PID control ensures that the reaming vane 1420 is precisely opened to the position corresponding to a 150mm reaming diameter. In the reaming section at a depth of 5m to 10m, the reaming vane 1420 cuts the hole wall, the drilling torque increases to 350~400 N·m, and the feed rate automatically decreases to 30mm / min.

[0096] After drilling to a depth of 10m, the reaming mechanism retracts, the telescopic mechanism 1200 retracts, and the drill bit withdraws from the borehole. The ultrasonic detector 1530 scans and images the borehole, showing a fractured zone radius of 0.3m and a plastic zone radius of 0.8m. The microseismic sensor 1540 collects two microseismic events, and inversion reveals that the peak stress around the borehole is located 0.6m from the borehole center, and the peak stress has decreased to 65% of the original rock stress.

[0097] The evaluation feedback module 4000 generates a pressure relief effect evaluation report: stress transfer distance 0.6m, stress release rate 35%, pressure relief effect is good. The evaluation results are fed back to the main control decision module 2000, and subsequent operation points will continue to use a 150mm enlargement diameter.

[0098] After the machine completes all the work points in sequence, it automatically returns to the starting position. The hydraulic support system 1020 retracts, the rotating mechanism retracts the drill rod into the vehicle body, and the tracked walking mechanism 1010 drives the equipment out of the tunnel, completing all the work tasks.

[0099] This invention is not limited to the above-described embodiments. Anyone can derive other products in various forms under the guidance of this invention. However, regardless of any changes in shape or structure, any technical solution that is the same as or similar to this application falls within the protection scope of this invention.

Claims

1. A method for autonomous operation control of a mining intelligent depressurization machine, characterized in that, Includes the following steps: Real-time image data of the surrounding rock in the tunnel is acquired, and the image data is inferred based on the YOLOv11 model to identify high-risk areas and their corresponding spatial location information. By integrating lidar point cloud data and visual image data, a three-dimensional semantic map of the alleyway is constructed, and an autonomous walking path is planned based on the spatial location information of the high-risk area. The pressure relief machine is controlled to move along the autonomous walking path to the target work point, and the drill rod rotation mechanism and telescopic mechanism are controlled to perform drilling operations according to the preset pressure relief parameters. When drilling reaches the preset depth, the hole reaming command is automatically triggered, controlling the umbrella-type hydraulic hole reaming mechanism to open radially and perform variable diameter hole reaming operation; Acquire ultrasonic imaging data and surrounding rock microseismic monitoring data inside the borehole, invert the stress distribution around the borehole, and generate an evaluation result of the pressure relief effect.

2. The autonomous operation control method for the intelligent depressurization machine in mining according to claim 1, characterized in that, The training dataset of the YOLOv11 model includes roadways with different cross-sectional shapes, different support methods, and on-site images before and after rockbursts. The model output includes the identification results of support damage areas, rock mass detachment areas, and structural instability areas.

3. The autonomous operation control method for the intelligent depressurization machine in mining according to claim 1, characterized in that, The construction of a three-dimensional semantic map of a tunnel specifically includes: generating a three-dimensional point cloud map using a laser SLAM algorithm, extracting semantic features using a visual SLAM algorithm, and fusing semantic labels with the three-dimensional point cloud to form a semantic map that includes support structures and hazardous area markers.

4. The autonomous operation control method for the intelligent depressurization machine used in mining according to claim 1, characterized in that, The autonomous walking path is planned using Dijkstra's algorithm, which takes the current location of the equipment as the starting point, the target work point as the ending point, and the passage area of ​​the alley as the constraint to generate the globally optimal path. During the walking process, a local obstacle avoidance algorithm is used to avoid sudden obstacles.

5. The autonomous operation control method for the intelligent depressurization machine in mining according to claim 1, characterized in that, The variable diameter reaming operation includes: maintaining the conventional borehole diameter in the shallow surrounding rock section, and controlling the umbrella-type hydraulic reaming mechanism to open step by step according to the preset diameter value after drilling to the preset depth, so as to realize deep reaming. The preset diameter value is related to the roadway support parameters and the stress inversion results of the surrounding rock.

6. The autonomous operation control method for the intelligent depressurization machine used in mining according to claim 1, characterized in that, The timing of the hole enlargement command is dynamically adjusted based on the drilling depth, drill rod torque changes, and surrounding rock hardness. When the drilling depth reaches a set threshold and the torque change meets preset conditions, the hole enlargement action is automatically executed.

7. The autonomous operation control method for the intelligent depressurization machine in mining according to claim 1, characterized in that, The inversion of the stress distribution around the borehole includes: identifying the range of the fractured and plastic zones within the borehole based on ultrasonic imaging data; combining the waveform characteristics and energy release parameters in the microseismic monitoring data; calculating the stress field distribution around the borehole using a stress inversion model; and comparing it with the theoretical stress distribution model.

8. The autonomous operation control method for the intelligent depressurization machine in mining according to claim 1, characterized in that, It also includes safety interlock control steps: before drilling operations, the locking status of the hydraulic support system, the positioning status of the drill rod, and the ambient gas concentration are automatically checked. Drilling operations are prohibited if any condition is not met. During the operation, if the concentration of explosive gas exceeds the limit, the power supply to the non-intrinsically safe circuit is immediately cut off and the equipment is automatically controlled to retreat to a safe area.

9. An autonomous operation control system for a mining intelligent pressure relief machine, characterized in that, include: The sensing module includes a high-definition camera, lidar, ultrasonic detector and micro-vibration sensor, used to collect tunnel images, point cloud data, borehole imaging and surrounding rock vibration data; The main control decision module, equipped with the YOLOv11 model and ROS2 architecture, is used to identify high-risk areas, build semantic maps, plan autonomous walking paths, and generate drilling and reaming control commands. The execution control module includes drivers for the drill rod rotation mechanism, the telescopic mechanism, and the umbrella-type hydraulic reaming mechanism, used to receive and execute the control commands; The evaluation feedback module is used to invert the stress distribution around the hole and generate an evaluation result of the pressure relief effect, which is then fed back to the main control decision module to optimize subsequent operation parameters.

10. The autonomous operation control system of the intelligent mine pressure relief machine according to claim 9, characterized in that, The main control decision module includes an intrinsically safe industrial computer for mining and a lower-level real-time controller. The industrial computer runs the ROS2 architecture and undertakes visual recognition and path planning algorithm calculations. The lower-level real-time controller communicates with the execution control module through the CANopen bus to realize real-time closed-loop control of drill pipe feeding, hole reaming action and travel drive.

Images