工矿自动化

Research progress on intelligent coal caving theory and technology

WANG Jiachen, YANG Shengli, LI Lianghui, ZHANG Jinwang, WEI Weijie

2024, 50(9): 1-12. doi: 10.13272/j.issn.1671-251x.18213

<Abstract>(174) <HTML> (24) <PDF>(56)

Abstract:
The longwall top-coal caving technology is an effective method for extracting thick and ultra-thick coal seams, and it has become a hallmark technology in China's coal mining industry. This paper reviews the research progress on the "Four elements" coal caving theory, the relationship between the top coal recovery rate and the rock mixed ratio, a recovery rate prediction model based on block distribution, and the relationship between instantaneous rock mixed ratio and cumulative rock mixed ratio. The challenges of intelligent coal caving technology are analyzed, emphasizing that the rock mixed ratio is a key factor affecting the top coal recovery rate and coal quality. Rapid and accurate calculation of the rock mixed ratio during the coal caving process is crucial for breakthroughs in intelligent coal caving technology. This technology is categorized into two types: non-image recognition and image recognition. The research progress, advantages, disadvantages, and usage conditions of different technologies are discussed in detail. Non-image recognition intelligent coal caving technology includes memory coal caving technology, sound and vibration signal detection technology, γ-ray detection technology, ground penetrating radar technology, microwave irradiation combined with infrared detection technology, and laser scanning coal caving monitoring technology. Image-based intelligent coal caving technology encompasses precise control of underground illumination environment, dust removal algorithms for coal caving images, accuracy assurance strategies for rock mixed ratio calculations, and infrared image recognition of coal and rock.

Study on emissivity measurement of different types of coal and gangue using the matching method

ZHANG Jinwang, HE Geng, HAN Xing, ZHANG Jiaming

2024, 50(9): 13-19, 27. doi: 10.13272/j.issn.1671-251x.2024070055

<Abstract>(112) <HTML> (27) <PDF>(9)

Abstract:
The type, surface texture, metamorphic degree, and developmental stage of coal and gangue significantly influence their emissivity. Accurate settings for emissivity parameters are essential for infrared temperature measurements and the identification of coal and gangue in infrared images. This study proposed a method for measuring the emissivity of coal and gangue based on the matching method. The approach integrated surface thermocouples with infrared thermography to assess emissivity. Samples were uniformly heated in a closed electric furnace, and once the temperature stabilized, a surface thermocouple measured the actual temperature of a selected area (denoted as t₁). Concurrently, the infrared thermography system measured the temperature of the same area (denoted as t₂). The emissivity of the infrared thermography system was calibrated until t₂ equaled t₁. At this point, the calculated emissivity reflected the true emissivity of the coal and gangue at that temperature. The experimental results indicated that: ① Under isothermal conditions, greater surface roughness of coal and gangue correlated with higher emissivity values, suggesting that surface roughness is a fundamental factor restricting the emissivity of these materials. ② The emissivity of four different types of coal and gangue decreased with increasing temperature, following a power function, with the fitting function's correlation coefficient (R²) exceeding 0.98, thereby confirming the feasibility of the matching method for measuring emissivity. ③ The inverse method revealed that the error rates between the measured and theoretical values under varying temperature conditions were all below 3%, validating the accuracy of the measured emissivity of coal and gangue.

Current status and prospects of surrounding rock control and intelligent coal drawing technology in fully mechanized caving face

PANG Yihui, GUAN Shufang, JIANG Zhigang, BAI Yun, LI Peng

2024, 50(9): 20-27. doi: 10.13272/j.issn.1671-251x.18211

<Abstract>(80) <HTML> (14) <PDF>(23)

Abstract:
This paper analyzes the current status and existing issues in the control technology of surrounding rock and intelligent top coal caving technology for thick and ultra-thick coal seams in fully mechanized caving faces. The study focuses on five aspects: efficient support of roadway surrounding rock, advanced support of working faces, the caving behavior of hard ultra-thick top coal, hydraulic support position monitoring, and intelligent top coal caving. To tackle the technical challenges and engineering demands for safe, efficient, and intelligent caving mining, research was conducted on surrounding rock control technology and intelligent coal caving technology. A mechanical model for cantilever beams of hard ultra-thick top coal was developed, and key technologies to enhance caving characteristics and extraction rate of top coal were created, facilitating large-height caving mining of hard ultra-thick coal seams. A modular advanced hydraulic support with a rotating self-resetting device was developed, allowing the hydraulic support's beam to automatically rotate based on the inclination angle of the roadway roof, significantly improving its adaptability to the roof and floor of roadway. The idea of replacing traditional bolt-mesh support with hydraulic supports for roadway support was proposed, offering high support efficiency, low cost, and savings on advanced support. A monitoring device and algorithm for the support posture of fully mechanized caving hydraulic supports based on the stroke of the jacks of columns and tail beams were developed, enhancing calculation efficiency and accuracy. An intelligent coal drawing control method integrating transparent geological models, coal volume monitoring devices, and coal and gangue identification devices was proposed, effectively addressing the challenges of intelligent coal drawing from ultra-thick top coal with multi-gangue layers. The paper concludes that trends in intelligent fully mechanized caving mining technology and equipment include intelligent geological assurance technology, precise measurement and intelligent sensing via machine vision, adaptive control technology for fully mechanized caving mining equipment, and digital twin technology.

Study on precise control of coal caving mechanisms based on the kinematics of support structures

WANG Zuguang, WANG Shen, LI Dongyin, LI Huamin, WANG Wen, YUE Shuaishuai, LI Donghui

2024, 50(9): 28-40. doi: 10.13272/j.issn.1671-251x.18212

<Abstract>(60) <HTML> (21) <PDF>(10)

Abstract:
Precise control of the coal caving mechanism is a crucial foundation for realizing intelligent and unmanned top coal caving mining. The spatial relationship between the coal caving mechanism and the rear scraper conveyor, as well as the top coal influence of the hydraulic support's posture on this spatial relationship, is key to constructing a control model for the caving support. Using the ZF17000/27.5/42D low-position top coal caving support as the research object, this study explained the spatial relationship between the coal caving mechanism and the rear scraper conveyor under different pitch angles of the support's roof and base. Based on the control logic for the opening degree of the hydraulic support's coal caving mechanism, a support posture sensing system was established, and a method for kinematic analysis of the hydraulic support's coal caving mechanism was proposed. A kinematic model for the end of the low-position hydraulic support's coal caving mechanism based on the D-H matrix was developed, and a calculation model for the opening degree of the hydraulic support's coal caving mechanism was constructed. The average calculation error was only 1.71%, meeting the accuracy requirements for field applications. A closed-loop control method for the coal caving mechanism based on posture feedback was proposed, and the coal caving decision model developed from the opening degree calculation model was applied in the field. Application results showed that during automatic coal caving, the mean square deviation of the average caving time for each support was only 0.13 minutes, with an overall caving efficiency improvement of 20%-43.9% compared to manual caving. The top coal recovery rate reached 89%, and the load on the rear scraper conveyor was more balanced, with an overload rate of only 0.73%.

Planning coal drawing control system based on process engine

YAO Yupeng, SHANG Chuhao, LIU Qing

2024, 50(9): 41-46, 107. doi: 10.13272/j.issn.1671-251x.2024030041

<Abstract>(83) <HTML> (22) <PDF>(14)

Abstract:
Current research on intelligent fully mechanized coal caving mining primarily focuses on perception, with limited studies on the intelligence of the coal drawing process. Existing automatic coal drawing control technologies face issues such as insufficient adaptability, low efficiency, and difficulty in quality control. To enhance the intelligence and operational efficiency of the coal drawing process, a planning coal drawing control system based on a process engine was designed. This system consisted of a coal drawing management unit and a window decision-making unit. The planning coal drawing management unit employed an asynchronous progressive scheduling strategy, flexible switching technology, and a process editing engine to achieve automated sequential coal drawing with weak correlation to the mining machine's position and online process editing. By associating with the load of the rear scraper conveyor, the system dynamically adjusted process starts and stops, ensuring safe operation of the scraper conveyor. The window decision-making unit utilized a PID control algorithm to dynamically adjust the tail beam angle, implementing feedback control of the coal drawing window. A genetic algorithm optimized a BP neural network to make intelligent decision about the size of the coal drawing window to adapt to varying operating conditions and improve coal drawing quality. Field application results indicated that the asynchronous progressive scheduling strategy and flexible switching technology enhanced the efficiency of automatic operation, eliminating the need for manual intervention. The number of automated operations per shift increased by 33.3%. The system's associated rear scraper conveyor load, pump station, and other equipment could dynamically adjust process starts and stops, resulting in a 61.1% decrease in the average stopping frequency of the rear scraper conveyor per shift, ensuring operational safety. The process editing engine accommodated various applications, substantially reducing adjustment time. The overlap of rear and front actions shortened the average operation time by 9.3%, increasing extraction efficiency. The correlation control of the tilt angle and intelligent decision-making for the planning coal release window improved daily calorific value by 10.3%, enhancing coal drawing quality.

Simulation study of top coal caving and conveying process based on smoothed particle hydrodynamics

LIU Bo, ZHANG Qiang, LIU Yang, DONG Xiangwei

2024, 50(9): 47-58. doi: 10.13272/j.issn.1671-251x.2024060003

<Abstract>(82) <HTML> (19) <PDF>(7)

Abstract:
Currently, in the numerical simulation research on the release laws of top coal during fully mechanized mining, complex coupling algorithms are required to address the continuity-discontinuity issues of top coal movement and ensure precise interaction of coal-rock interface information. However, the conveying process of scraper conveyors is typically neglected in these simulations. To address this problem, a meshless numerical computation model was constructed based on smoothed particle hydrodynamics (SPH). The discrete equations of SPH, derived from the control equations of continuous medium mechanics, were established. An elastic-plastic soil constitutive model along with the Drucker-Prager yield criterion were introduced to achieve dynamic simulation of the caving, movement, and release processes of the top coal. Considering the actual coal release and conveying processes in the mining area, a scraper conveyor model was constructed to simulate the release of top coal and the conveying of bottom coal along the working face, obtaining the variations in coal-rock interface and coal flow velocity at different scraper conveyor operating speeds (0-1.5 m/s). The simulation results indicated that the elastic-plastic soil constitutive model effectively simulated the flow behavior of particles. By setting parameters such as friction angle and elastic modulus, the issue of uncertain parameters commonly found in traditional discrete element models was avoided. After stabilization of the coal flow velocity, the stress distribution of the top coal near the coal drawing outlet exhibited a "double peak" pattern. The operating speed of the scraper conveyor significantly impacted the coal drawing time, while its effect on the coal-rock interface at termination and the shape of the released body was minimal. When multiple supports released coal simultaneously, the conveying capacity of the scraper conveyor needed to be considered, as interference in bottom coal transportation between different supports could lead to blockage effects at the release port. The "gangue closing" rule resulted in variations in the amount of coal drawing at different coal drawing outlets, with the standard deviation of top coal drawing amount from 40 coal drawing outlets (7.52 m²) being greater than that of automatic coal drawing (1.93 m²).

Detection method for gangue mixed ratio in fully mechanized caving faces based on the GSL-YOLO model

WANG Fuqi, WANG Zhifeng, JIN Jiancheng, JING Qinghe, WANG Yaohui, WANG Dalong, WANG Yilong

2024, 50(9): 59-65, 137. doi: 10.13272/j.issn.1671-251x.2024080011

<Abstract>(83) <HTML> (20) <PDF>(19)

Abstract:
Aiming to address the issues with current gangue mixed ratio detection methods in fully mechanized caving face based on deep learning, such as low accuracy of coal gangue identification, poor segmentation performance, large model parameters and computation load, and the inability to achieve real-time detection of gangue mixed ratio under complex conditions such as low lighting, high dust, and coal and angue stacking, the paper proposed a gangue mixed ratio detection method based on the GSL-YOLO model. The GSL-YOLO model introduced the following improvements to the YOLOv8-seg model: the incorporation of a global attention mechanism (GAM) in the backbone network to enhance feature extraction by reducing information dispersion and amplifying global interaction representation; the use of a spatial pyramid pooling with efficient local aggregation network (SPPELAN) module to improve detection performance for targets of varying sizes; and the adoption of a lightweight asymmetric dual-head (LADH) to reduce training difficulty while increasing inference speed. Additionally, a gangue mixed ratio calculation method based on category segmentation masks was proposed, which calculated the ratio of the pixel area of gangue to the total pixel area in the segmentation mask of coal flow images, serving as the instantaneous gangue mixed ratio. Experimental results showed that: ① The GSL-YOLO model achieved an mAP@0.5∶0.95 of 96.1%, which was 0.8% higher than the YOLOv8-seg model. ② The GSL-YOLO model had 2.9×10⁶ parameters, 11.4×10⁹ floating-point operations, and a model weight of 6.0 MiB, representing reductions of 12.1%, 5.8%, and 11.8% respectively compared to the YOLOv8-seg model, achieving model lightweighting. ③ The GSL-YOLO model achieved a frame rate of 12 frames per second on the test set, essentially meeting the requirements for real-time detection. ④ Compared with the YOLO series models, the GSL-YOLO model had the best segmentation effect, the highest detection accuracy, fewer parameters and computation load, and the best overall performance. ⑤ Based on three frames of images captured from the coal flow on the rear scraper conveyor of the fully mechanized caving face, the instantaneous gangue mixed ratio was calculated, and the results showed that the proposed method successfully realized real-time calculation of the gangue mixed ratio in fully mechanized caving face.

Research on the characteristics of the falling behavior of mixed coal and gangue

SHAN Pengfei, YANG Tong, SUN Haoqiang, XI Bojia

2024, 50(9): 66-74. doi: 10.13272/j.issn.1671-251x.2024070058

<Abstract>(43) <HTML> (14) <PDF>(4)

Abstract:
Traditional studies on the dynamic characteristics of coal flow during the top coal caving process, based on image detection technology, have primarily focused on specific-stage image analysis, lacking a comprehensive analysis of dynamic characteristics across all stages. Existing research has rarely integrated the changes in the loose zone of the overlying strata with coal and gangue separation and coal flow characteristics during top coal caving, resulting in a lack of systematic and holistic analysis of the entire coal caving process. In response to these issues, this study systematically investigated coal flow dynamics, coal and gangue separation effectiveness, and the subsidence of the loose zone in the overlying strata during the top coal caving. First, this paper proposed a dynamic analysis method for the top coal caving process based on a dual optical flow network. The results indicated that the coal caving speed was not affected by the caving method and pattern, and that average detection accuracy increased with the number of caving openings, exhibiting a notably linear increase during the periodic caving stage. The release rate of top coal showed a positive correlation with average detection accuracy, validating the effectiveness of the method in the top coal caving process monitoring. Second, OpenCV technology was used to conduct experimental analysis on the subsidence area of the loose zone in the overlying strata. Results demonstrated that the subsidence area grew sharply during the initial caving stage and gradually stabilized over time. The dynamic changes in the subsidence area effectively indicated the progression of top coal release, enabling transparent monitoring of release process. Finally, based on data from weighing experiments, the relationships among caving amount, release rate, and gangue content were analyzed. Results showed that the amount of pure coal release was the highest in the initial caving stage and stabilized in the periodic caving stage, while gangue content decreased as the number of caving openings increased. These findings further reveal the influence of caving methods on coal and gangue separation and the release rate of top coal.

Experimental study on the evolution characteristics of dynamic load of hydraulic support top beam during coal caving

HUO Yuming, HU Wenshuo, GAO Peng, YAN Chuan

2024, 50(9): 75-81. doi: 10.13272/j.issn.1671-251x.2024080001

<Abstract>(65) <HTML> (15) <PDF>(6)

Abstract:
Contact coal-gangue identification requires studying the bearing characteristics of hydraulic support top beams in fully mechanized top coal caving. However, existing research primarily focuses on the bearing characteristics of supports before and after coal caving or the mechanical response characteristics of supports under given loads, neglecting an in-depth exploration of load changes during the coal caving process. To address this issue, a dynamic load similarity simulation test platform for top-coal hydraulic supports was established, using granular particles to simulate broken coal gangue. This setup simulated the coal caving process in a fully mechanized working face, and thin-film pressure sensors were employed to collect pressure data from the support top beams. The dynamic load evolution characteristics of the support top beams during the coal caving process were analyzed. The experimental results indicated: ① The caving of top coal significantly affected the load on the support top beams, demonstrating an evolution pattern where the overall load first increased, then decreased, and finally stabilized as the top coal was released. ② Along the length of the beam, the locations of the support top beams farther from the protective beam were less affected by the caving of top coal. This was primarily reflected in the smaller increase in peak load compared to the initial value at locations farther from the protective beam, as well as a longer time required to reach the peak load. ③ Along the width of the beam, due to the constraints of boundary conditions or the unevenness of the flow process during top coal caving, the peak load at different positions of the beam showed variability, with the maximum increase in peak load compared to the initial value reaching up to 2.4 times the minimum increase.

Research on filling, pressure relief, and rock burst prevention in horizontal sublevel fully mechanized top coal caving of near-vertical coal seam groups

WANG Ziwei, CHENG Boyuan, WEI Weijie, SUN Wenchao, XIE Dongheng, XIE Mianyu

2024, 50(9): 82-89. doi: 10.13272/j.issn.1671-251x.2024070119

<Abstract>(67) <HTML> (14) <PDF>(14)

Abstract:
Current rock burst prevention measures for near-vertical coal seam mining primarily include blasting, hydraulic fracturing, and the establishment of protective layers. These measures either damage interlayer rock pillars and roof/floor strata or prove inadequate in resolving stress concentration in interlayer rock pillars at large mining depths, often resulting in significant surface subsidence. Using the Wudong Coal Mine as the engineering context, this study proposed a filling technique for goafs in horizontal sublevel fully mechanized top coal caving of near-vertical coal seam groups. This technique was intended to support interlayer rock pillars and roof/floor strata, reducing stress concentration in the surrounding coal and rock masses of the mining segments. Three filling schemes were designed: Scheme 1 involved filling the goaf in the first mining segment with high-strength materials, with ordinary materials used in other segments; Scheme 2 involved filling the goaf in the first segment with high-strength materials, with alternating high-strength and ordinary materials in other segments; and Scheme 3 involved filling the goaf in each segment with high-strength materials. Numerical simulations were conducted to assess the pressure relief and rock burst prevention effectiveness of the three schemes. Results indicated that, compared to no filling, the maximum vertical stress in interlayer rock pillars decreased by 25.07%, 26.57%, and 29.23% under the three schemes, respectively, while the maximum horizontal stress in the coal body of the subsequent segment decreased by 10.63%, 10.79%, and 12.34%, respectively. Considering both pressure relief effectiveness and economic feasibility, Scheme 3 with interval filling was identified as the optimal solution. It was suggested that this approach be combined with real-time intelligent monitoring technology in high-stress areas to promptly support interlayer rock pillars, thus reducing stress concentration and preventing rock bursts.

Research on the optimal position of roadways in fully mechanized caving faces in mine-out areas of close distance coal seams

ZHANG Wei, ZHANG Guojun, SHI Yongguang, ZHEN Weijie, WANG Yuliang, LI Yihang, LI Yang

2024, 50(9): 90-97. doi: 10.13272/j.issn.1671-251x.2024070074

<Abstract>(88) <HTML> (31) <PDF>(8)

Abstract:
Fully mechanized caving faces in close-distance coal seams involve extensive extraction spaces and high mining intensity. The extraction of roadways in lower coal seams is affected by stress concentration and support challenges resulting from the mining of upper seams. Hence, determining the optimal roadway position is crucial for effective support control in these settings. This study focused on the No. 2 coal seam and the No. 1-1 sub-seam at Xilutian Coal Mine. It evaluated both the stress reduction zone in the floor caused by upper seam extraction and the limit equilibrium zone during lower seam extraction, concluding that the optimal roadway position should be more than 22.79 meters away from the solid coal pillar. Theoretical calculations were used to analyze the stress distribution pattern in the floor following upper seam extraction, as well as the deformation and failure characteristics of the surrounding rock at various internal offsets. The results revealed: ① A pronounced difference between maximum and minimum stresses occurred closer to the floor of the mine-out area. ② With increasing internal offset, the surrounding rock stress and stress concentration coefficient initially decreased sharply, then increased slowly, and eventually stabilized, with relatively low values observed within the 20-25 meters internal offset range. ③ The plastic zone of the surrounding rock decreased and then increased, with minimal damage to the roadway rock observed at internal offsets of 20 and 25 meters. ④ Roadway deformation decreased as the internal offset increased, and surrounding rock displacement stabilized when the internal offset reached 25 meters. ⑤ The optimal internal offset for the roadway was determined to be 20-25 meters. Engineering applications confirmed that a 24-meter internal offset maintained both rock looseness and deformation within controllable limits, further validating this internal offset.

Study on the spatiotemporal distribution of coal flow in the scraper conveyor of fully mechanized mining face

CHEN Shuhang, WANG Shibo, GE Shirong, WANG Yun, MA Guangjun

2024, 50(9): 98-107. doi: 10.13272/j.issn.1671-251x.2023110009

<Abstract>(103) <HTML> (33) <PDF>(20)

Abstract:
Research on coal flow characteristics based on sensors is limited by the restricted monitoring range of sensors, making it impossible to study the coal flow characteristics of the entire scraper conveyor. Additionally, research on coal flow characteristics based on model simulations often lacks consideration of mining processes, preventing the prediction of the spatiotemporal distribution of coal flow across the entire scraper conveyor. To address the issue of difficulty in monitoring the coal flow characteristics of the entire scraper conveyor in a fully mechanized mining face, this study integrated the mining process of the fully mechanized face. By analyzing the processes of coal cutting and loading by the shearer and the coal transportation by the scraper conveyor, a mathematical model for the instantaneous loading volume and cross-sectional area of the scraper conveyor under different loading methods in various process segments was established. The coal flow transportation process of the scraper conveyor was divided into coal flow translation and loaded coal flow superposition, and a spatiotemporal distribution prediction model for coal flow on the fully mechanized face scraper conveyor was developed based on the finite element method. Using this model, the spatiotemporal distribution characteristics of coal flow on the scraper conveyor during the mining process cycle were analyzed through simulation. Compared to the normal cutting stage in the middle, the spatiotemporal distribution of coal flow was more complex during the cutting stage at the ends. The maximum cross-sectional area of the loaded coal flow in the middle trough occurred during the stage of drum swapping. The volume of coal flow transported by the scraper conveyor changed in opposite trends during the upward and downward movements of the shearer, with the trend determined by the shearer's traction direction. Actual operating data from a shearer and scraper conveyor in a mine were used as input parameters for the model, and the coal volume was calculated based on the predicted spatiotemporal distribution. The results showed that the predicted trend of coal volume was consistent with on-site measurements, with a cumulative coal volume prediction error of 9.24%. The coal volume prediction errors during the fixed time periods of the shearer's cutting process and upward movement stage were 13.19% and 13.78%, respectively, demonstrating the accuracy of the spatiotemporal distribution prediction model for coal flow.

Detection of surface defects on conveyor belts based on adversarial repair networks

YANG Zelin, YANG Liqing, HAO Bin

2024, 50(9): 108-114, 166. doi: 10.13272/j.issn.1671-251x.2024030002

<Abstract>(40) <HTML> (14) <PDF>(3)

Abstract:
In response to the challenges of acquiring and labeling defect data on conveyor belts, as well as the low accuracy of deep learning-based conveyor belt defect detection methods due to unstable factors and data fluctuations in working environments, this study proposed a surface defect detection model based on adversarial repair networks. The model primarily consisted of a generator with an autoencoder structure and a Markov discriminator. During the training phase, simulated surface defect images of the conveyor belt were input into the generator to obtain reconstructed images without simulated defects, enhancing the model's ability to generalize to unknown defects. The original undamaged conveyor belt images, reconstructed images, and simulated surface defect images were input into the Markov discriminator, and feature maps were obtained through a residual network, improving the model's detection capability for subtle defects. In the detection phase, the test image was input into the trained generator to obtain the reconstructed image, and the trained Markov discriminator was used to extract feature maps from both the test image and the reconstructed image. The anomaly score was calculated based on the mean squared error between the feature maps of the test image and the reconstructed image, as well as the maximum value of the feature map of the test image, and compared with a set threshold to determine whether the test image contained defects. Experimental results showed that the area under the receiver operating characteristic curve (ROC-AUC) of this model reached 0.999, the area under the precision-recall curve (PR-AUC) reached 0.997, and the detection time for a single image was 13.51 ms, which could accurately locate the positions of different types of defects.

LiDAR-based edge extraction method for underground belt conveyors

HUANG Chenxuan, CHANG Jian, WANG Lei

2024, 50(9): 115-123. doi: 10.13272/j.issn.1671-251x.2024060025

<Abstract>(94) <HTML> (28) <PDF>(4)

Abstract:
The belt conveyor is one of the inspection targets of the inspection robot in the unstructured belt roadway of underground coal mines. Extracting its edges allows the robot to obtain its spatial pose relative to the inspection target, providing environmental information to support the execution of inspection tasks. Currently, most underground edge extraction techniques are vision-based, which struggle to overcome challenges such as low illumination, heavy dust, and dense fog. To address this issue, an explosion-proof 16-line LiDAR was used as the sensor for the inspection robot to acquire the roadway point cloud, reducing the environmental impact on the extraction results. The raw sparse point cloud was preprocessed using statistical outlier removal and passthrough filtering to eliminate noise and irrelevant points. The belt conveyor's point cloud plane was segmented using the Random Sample Consensus (RANSAC) algorithm, and the edge point cloud of the belt conveyor was extracted using a projection-quad tree method. The combined rviz and Gazebo simulation results showed that, under different operating conditions of the robot, the accuracy of belt conveyor edge extraction was no less than 96.33%. When the LiDAR shielding rate was below 30%, the accuracy was no less than 79.23%. Laboratory tests showed that, even when the surface of the belt conveyor had a 100% water distribution and saturated thickness, the edge extraction accuracy was no less than 88%. Overall, this method outperforms the latitude and longitude extremum search method, the curvature threshold method based on KDTree/OcTree, and the adjacent point angle threshold method based on KDTree/OcTree, with an average computation time of only 36 ms, meeting the real-time inspection needs of underground environments.

Detection of underground personnel safety helmet wearing based on improved YOLOv8n

WANG Qi, XIA Lufei, CHEN Tianming, HAN Hongyin, WANG Liang

2024, 50(9): 124-129. doi: 10.13272/j.issn.1671-251x.2024040054

<Abstract>(162) <HTML> (39) <PDF>(21)

Abstract:
Existing methods for detecting safety helmet wearing among underground personnel fail to consider factors such as occlusion, small target size, and background interference, leading to poor detection accuracy and insufficient model lightweighting. This paper proposed an improved YOLOv8n model applied to safety helmet wearing detection in underground. A P2 small target detection layer was added to the neck network to enhance the model's ability to detect small targets and better capture details of safety helmets. A convolutional block attention module (CBAM) was integrated into the backbone network to extract key image features and reduce background interference. The CIoU loss function was replaced with the WIoU loss function to improve the model's localization capability for detection targets. A lightweight shared convolution detection head (LSCD) was used to reduce model complexity through parameter sharing, and normalization layers in convolutions were replaced with group normalization (GN) to reduce model weight while maintaining accuracy as much as possible. The experimental results showed that compared to the YOLOv8n model, the improved YOLOv8n model increased the mean average precision at an intersection over union threshold of 0.5 (mAP@50) by 1.8%, reduced parameter count by 23.8%, lowered computational load by 10.4%, and decreased model size by 17.2%. The improved YOLOv8n model outperformed SSD, YOLOv3-tiny, YOLOv5n, YOLOv7, and YOLOv8n in detection accuracy, with a complexity only slightly higher than YOLOv5n, effectively balancing detection accuracy and complexity. In complex underground scenarios, the improved YOLOv8n model were able to achieve accurate detection of safety helmet wearing among underground personnel, addressing the issue of missed detections.

Personnel localization method for low-visibility environments based on improved YOLOv3

LU Xiaoya, LI Haifang

2024, 50(9): 130-137. doi: 10.13272/j.issn.1671-251x.2024070085

<Abstract>(76) <HTML> (28) <PDF>(3)

Abstract:
In coal mines, inadequate lighting and dust obstruction result in personnel targets captured by video monitoring systems appearing as small or low-visibility objects in two-dimensional images. The original YOLOv3 network's Darknet53 feature pyramid structure was insufficient for effectively extracting and preserving detailed information about these targets, leading to inaccurate localization. To address this issue, personnel localization method for low-visibility environments based on improved YOLOv3 was. First, the clarity of coal mine monitoring videos under low-visibility conditions was enhanced using a combination of β function mapping and inter-frame information enhancement techniques. Next, Darknet53 in YOLOv3 was replaced with the lighter Darknet-19, and CIoU was introduced as the loss function to optimize personnel target identification in the enhanced video. Finally, the identified targets were projected from two-dimensional space to three-dimensional space based on the mapping model, completing the personnel localization process. Experiments conducted on monitoring videos from a coal mine in low-visibility conditions revealed the following findings: ① After applying the improved YOLOv3, the brightness, visibility, and various evaluation metrics (average gray level, average contrast, information entropy, and gray spectral bandwidth) of the video frames demonstrated significant improvements compared to the original videos. There was a substantial enhancement in overall lighting conditions and contrast, facilitating better differentiation between targets and backgrounds, thereby validating the effectiveness of the image enhancement techniques employed. ② The improved YOLOv3 accurately identified personnel in the video frames, with no instances of missed detections. ③ Using calibrated objects or manually annotated real three-dimensional positions as benchmarks, the deviation between the projected results and the actual positions was calculated (covering distance deviations in the X, Y, and Z directions). The deviations in both the X and Y directions were below 0.2 m, while the deviation in the Z direction was below 0.002 m, indicating a high mapping effect and localization accuracy of the constructed mapping model.

Automatic reasoning technology for coal mine industrial data AI models

ZHANG Zhixing, FU Xiang, ZHANG Xiaoqiang, LI Haojie, QIN Yifan, LIU Meng, SUN Yan, JIA Yifan, YANG Yuqi

2024, 50(9): 138-143. doi: 10.13272/j.issn.1671-251x.18181

<Abstract>(52) <HTML> (19) <PDF>(8)

Abstract:
The automation of coal mine production processes has largely relied on artificial intelligence (AI) technology to analyze industrial data. However, AI models developed for single application scenarios prove inadequate for the complex environments in coal mining. Relying solely on distributed computing to process the input features of AI models has led to decreased application efficiency. To address these challenges, an automatic reasoning technology for AI models in coal mine industrial data was developed. The system architecture consisted of three layers: the data layer, the computation-driving layer, and the model reasoning layer. The data layer gathered and stored various types of monitoring data, supplying raw data to the computation-driving layer. The computation-driving layer converted this vast amount of raw data into input features for AI models tailored to coal mining applications. An automatic switching mechanism between two computational engines—based on the input feature values—intelligently selected either Spark-based distributed computing or Python-based local computing, depending on the data volume, thereby resolving the issues of slow processing speeds and high latency in large-scale data applications. In the model reasoning layer, the input features were fed into the AI models for reasoning. A collaborative reasoning mechanism, with multiple triggering methods—scheduled, manual, and feedback-triggered—was introduced to enhance the effectiveness of AI models in complex coal mining scenarios. The results demonstrate that this technology enables rapid calculation of input features for multiple AI models across different application scenarios, as well as fast, automatic, and collaborative reasoning.

Intelligent identification of electromagnetic radiation signals induced by coal rock fractures using machine learning

LI Baolin, FENG Jiaqi, WANG Enyuan, SUN Xinyu, WANG Shuowei

2024, 50(9): 144-152. doi: 10.13272/j.issn.1671-251x.2024070019

<Abstract>(75) <HTML> (23) <PDF>(3)

Abstract:
Electromagnetic radiation (EMR) has proven to be an effective monitoring technology for coal rock dynamic disasters, including underground rock burst and coal and gas outbursts. However, the intricate generation mechanisms of electromagnetic signal, coupled with interference from underground environments, can compromise the accuracy of disaster monitoring and early warning systems. Accurately identifying EMR signals induced by coal rock fractures (effective signals) is essential for the widespread application of this technology. This study conducted monitoring experiments on electromagnetic radiation during uniaxial compression of coal rock, analyzing the time-domain, frequency-domain, and fractal characteristics of both valid and interference signals. Machine learning algorithms, such as linear discriminant analysis, support vector machines, and ensemble learning methods, were utilized to develop intelligent identification models for effective and interference signals. A comparative analysis of the recognition accuracy across different models was performed. The results demonstrated that characteristics like fractal box dimension, average frequency, count, and peak frequency effectively distinguished between valid and interference signals, with single-feature recognition accuracy surpassing 70%. Both the feature set and the choice of machine learning algorithm significantly influenced the identification accuracy of valid and interference signals. The ensemble learning method, leveraging the complete feature set, achieved the highest identification accuracy of 94.5% for both signal types, fulfilling the requirements for EMR monitoring and early warning applications.

Research on tetrahedral adaptive mesh grading refinement for intersecting faults

CHEN Yingxian, ZHU Zhe, MA Huiru, FU Jiepeng

2024, 50(9): 153-160. doi: 10.13272/j.issn.1671-251x.2024030058

<Abstract>(68) <HTML> (20) <PDF>(1)

Abstract:
Current tetrahedral adaptive mesh refinement techniques have primarily focused on the 3D reconstruction and analysis of simple stratified geological bodies. When applying adaptive mesh refinement to complex geological structures, such as those containing intersecting faults with discontinuous data, excessive refinement can easily lead to compromised mesh structures in the fault zones. To improve the accuracy of tetrahedral mesh models for such complex fault systems, this study proposed a tetrahedral adaptive mesh grading refinement method specifically for intersecting faults. Initially, the refinement range around the fault was adaptively determined based on a fault influence formula. Subdivision formulas were then developed for tetrahedrons and tetrahedral edges to grade both the tetrahedrons and their edges within the refinement range. To address the various scenarios that arose during tetrahedral mesh subdivision, the eight types of subdivisions were unified into three types by upgrading the edge treatments. Finally, new vertices were introduced, and existing vertices were reconnected to tetrahedrons within the refined area, adjusting mesh element sizes to generate a high-quality mesh model. A case study was conducted on a tetrahedral mesh model from an open-pit coal mine in Inner Mongolia. The mesh model was analyzed before and after refinement using a 3D mesh quality evaluation algorithm and FLAC3D simulation software. Results showed that the distortion value of the refined mesh model decreased from 0.3317 to 0.3061, indicating an improvement in mesh quality. Under the same parameters, the unrefined model exhibited a maximum displacement of 1.16 m with a stability coefficient of 1.27, while the refined model showed a maximum displacement of 1.29 m and a stability coefficient of 1.23. The displacement cloud map of the refined model was aligned with the fault, accurately reflecting the fault distribution and its impact on the slope. In contrast, the displacement cloud map of the unrefined model was misaligned with the fault center, demonstrating a less pronounced effect of the fault on the slope.

Safety analysis of lithium-ion batteries under mechanical shock conditions

NI Chunming

2024, 50(9): 161-166. doi: 10.13272/j.issn.1671-251x.2024050064

<Abstract>(96) <HTML> (28) <PDF>(4)

Abstract:
The harsh and confined environment of underground coal mines makes lithium-ion batteries vulnerable to external physical shocks or damage, potentially leading to safety incidents. This study investigated a 100 A·h lithium manganese oxide ion battery designed for mining applications, employing puncture, high-temperature, and humidity tests to evaluate its safety performance. The puncture test simulated mechanical shocks typical in coal mine environments by penetrating the battery with a sharp object and observing its response under extreme conditions. A furnace and humidity-controlled environment chamber were also utilized to replicate the high-temperature and humid conditions encountered in coal mines, assessing the battery's safety and reliability post-puncture. The results revealed the following: ① After puncture by a tungsten needle, the battery exhibited surface deformation and cracking, but no electrolyte leakage, smoke, fire, or explosion occurred, with no gas generation inside. Although the temperature of the punctured battery rose significantly, it remained within a safe range without igniting or exploding, indicating a certain level of thermal stability suitable for coal mine applications. ② The punctured battery expanded notably when heated in the furnace, accompanied by gas leakage; however, no explosion or combustion took place, suggesting thermal stability under specific conditions. ③ In a humid environment, the punctured battery produced gas, leading to increased internal pressure. The combination of puncture and humidity raised the battery's temperature, but the moisture acted as a cooling agent, resulting in a slower temperature increase compared to high-temperature conditions, without triggering explosion or combustion. This indicated that the battery maintained thermal stability in humid environments and did not exhibit thermal runaway.

2024 Vol. 50, No. 9

Submission System

Latest AcceptedMore>