Current Issue

2024 Vol. 50, No. 8

column
  Previous Issue
Display Method:
Cover Catalogue
Special of Intelligent Mining Control Technology for Mines
Theory and method of shearer digital twin navigation cutting motion planning
MIAO Bing, GE Shirong
2024, 50(8): 1-13. doi: 10.13272/j.issn.1671-251x.2024070063
<Abstract>(292) <HTML> (199) <PDF>(126)
Abstract:
In order to further improve the intelligence level of coal working faces and achieve autonomous deduction, autonomous learning, and autonomous optimization of shearer navigation cutting, based on the concept of shearer autonomous navigation cutting technology and digital twin smart mining face, the theory and method of shearer digital twin navigation cutting motion planning are proposed. It includes the theory of digital twin and the construction method of shearer digital twin navigation cutting motion planning system based on this theory. Based on the theory of digital twins, the paper explores the physical scenarios of smart mining working face, the construction of digital twin models, and the driving, interactive, and evolutionary mechanisms of digital twins. To meet different application needs, digital twin models are divided into physical entities, twin models, and twin data models. The features of these three types of models are analyzed in detail. Three operational mechanisms, including model driven, data-driven, and service driven, are introduced. The three operational mechanisms achieve the transition from perceptual intelligence to cognitive intelligence through virtual real interaction logic. The study develops a shearer digital twin navigation cutting motion planning system. The system supports the service functions of digital twin cutting status, dynamic navigation map, digital twin reinforcement learning environment, and reinforcement learning motion planning through physical perception layer, comprehensive data layer, data fusion analysis layer, and digital twin service layer. By digital means, the navigation and cutting process of the shearer in reality is replicated in the digital twin operating environment. The adaptive fusion, intelligent analysis, and optimal planning of data are achieved through the calling of various modules within the system. Finally, by comparing the performance of the deep q-network with normalized advantage functions(DQN-NAF) algorithm and the deep deterministic policy gradient (DDPG) algorithm in the motion planning task of shearers in the constructed digital twin environment, the results show that the DQN-NAF algorithm exhibits better performance and stability in solving the digital twin motion planning task of shearers.
Attitude monitoring method for hydraulic support in fully mechanized working face based on PSO-ELM
LI Lei, XU Chunyu, SONG Jiancheng, TIAN Muqin, SONG Danyang, ZHANG Jie, HAO Zhenjie, MA Rui
2024, 50(8): 14-19. doi: 10.13272/j.issn.1671-251x.2024070023
<Abstract>(69) <HTML> (20) <PDF>(2)
Abstract:
In response to the problems of cumulative errors and inaccurate correction results in the attitude calculation method of hydraulic supports based on inertial measurement units, a fully mechanized working face hydraulic support attitude monitoring method based on particle swarm optimization (PSO) - extreme learning machine (ELM) is proposed. Using the pitch angle of the hydraulic support top beam as the monitoring object, a tilt sensor and gyroscope are used to collect real-time information on the support attitude of the hydraulic support top beam. The collected data is preprocessed and input into the PSO-ELM error compensation model to obtain the predicted solution error. At the same time, the hydraulic support attitude is calculated through Kalman filtering fusion to obtain the calculated value. Then the method uses the error prediction value to compensate for the error in the calculated value, in order to obtain more accurate data on the top beam support attitude. This method only considers the relationship between acceleration and angular velocity data and solution errors, without relying on specific physical models. It can effectively reduce the cumulative error of attitude solution. The experimental results show that the average absolute error of the pitch angle of the top beam of the hydraulic support has been reduced from 1.420 8° before compensation to 0.058 0°. The error curve has good convergence, verifying that the proposed method can sustainably and stably monitor the support attitude of the hydraulic support.
Research on dynamic features of hydraulic system for fully mechanized mining support and its improvement design
GUO Xinwei
2024, 50(8): 20-29. doi: 10.13272/j.issn.1671-251x.2024060046
<Abstract>(77) <HTML> (24) <PDF>(9)
Abstract:
In the hydraulic support work of fully mechanized working face, there are often problems such as insufficient initial support force and slow moving speed of the support. Currently, most solutions are based on the steady-state operation law of the hydraulic system of the support, such as increasing the flow rate of the pump station and reducing pressure loss. There is relatively little research on the dynamic features of the hydraulic system. The dynamic equation of the hydraulic system for the fully mechanized mining support is established. The dynamic features of the hydraulic system related to the initial support force and moving speed of the support, as well as the hydraulic impact features of the emulsion pipeline system, are theoretically analyzed. It is found that the approximate no-load operation of the column or jack and the hydraulic impact of long-distance pipelines are the main reasons for the significant pressure drop and fluctuation in the hydraulic system of the support. The mechanism of hydraulic impact in the hydraulic system of the bracket is revealed to be the sudden opening and closing of the electro-hydraulic directional valve and the pressure of the column touching the top. The correctness of the theoretical analysis is verified through on-site measured data and AMESim simulation. An improvement plan for the hydraulic system of the fully mechanized mining support is proposed. It includes the installation of multiple accumulators on the support, the addition of hydraulic control one-way valves and electro-hydraulic directional valves to control the charging and discharging methods of the hydraulic system accumulators at different stages of frame movement. The method uses the instantaneous high flow features of the accumulators and the overpressure generated by the peak hydraulic impact pressure of long-distance pipelines to enhance the initial support force of the support. The simulation results show that the improved system can effectively increase the initial support force and frame moving speed of the hydraulic support.
Periodic pressure prediction of working face based on dynamic adaptive sailfish optimization BP neural network
YAO Yupeng, XIONG Wu
2024, 50(8): 30-37. doi: 10.13272/j.issn.1671-251x.2024060060
<Abstract>(69) <HTML> (16) <PDF>(3)
Abstract:
In order to solve the problems of insufficient precision, poor generalization, and high computational requirements of existing methods for periodic pressure prediction of working face, a periodic pressure prediction model of working face based on dynamic adaptive sailfish optimization BP neural network (DASFO-BP) is proposed. By analyzing the mechanism of working face periodic pressure, the influencing factors related to pressure are obtained. The Pearson correlation coefficient is used to determine the factors that have a significant impact on pressure (advance speed, direct roof thickness, basic roof thickness, mining height, coal seam dip angle, and dip length) as inputs for the prediction model. The subsequent pressure intensity and pressure step distance are used as outputs for the prediction model. A dynamic adaptive optimization strategy is proposed to improve the robustness of the sailfish optimization (SFO) algorithm. In the early stage of optimization, SFO is used to achieve fast convergence, while in the middle stage, bald eagle search (BES) is used to escape local optima. In the later stage, the advantage of particle swarm optimization (PSO) deep search is utilized to improve the precision of the solution. A dynamic adaptive sailfish optimization (DASFO) algorithm is improved to train the hyperparameters of the BP neural network, and a pressure prediction model based on DASFO-BP is constructed. The experimental results indicate that the DASFO algorithm can achieve fast convergence on both unimodal and multimodal test functions. Compared with BP, SFO-BP, and NCPSO-BP, DASFO-BP has higher precision in predicting the intensity and step distance of periodic pressure, and has strong generalization ability and fitting capability. It can accurately predict the pressure and its distribution in the next period.
Coal-rock image recognition method integrating drilling geological information
LI Ji, MA Xiaofeng, WU Jieqi, QIANG Xubo, WU Liyang, YAN Bo, DONG Jihui, CHEN Chaosen
2024, 50(8): 38-43, 68. doi: 10.13272/j.issn.1671-251x.2024040048
<Abstract>(86) <HTML> (20) <PDF>(5)
Abstract:
The current deep convolutional neural network models applied to coal-rock image recognition have problems such as large volume and cumbersome calculation process. It is difficult to meet real-time detection requirements, and it has poor adaptability to complex environments such as low lighting and high dust. In order to solve the above problems, a coal-rock image recognition method integrating drilling geological information is proposed. Firstly, the improved spectral residual saliency detection (ISRSD) algorithm is used to enhance the quality of coal-rock images, effectively reducing the adverse effects of complex environments on the features of coal-rock images. Secondly, the method uses the attentional VGG (AVGG) deep convolutional neural network model. The AVGG performs pruning based on VGG, adds convolutional block attention module (CBAM), and introduces adaptive learning rate adjustment strategy to efficiently extract coal-rock image features. Finally, the Bayesian model is used to integrate the features of coal-rock images with the geological information obtained from the borehole geological column chart, in order to improve the accuracy and robustness of coal-rock classification. The experimental results show that the image enhanced by the ISRSD algorithm has more prominent targets, lower color distortion, and relatively complete preservation of image features such as edges and textures. The accuracy of the AVGG model is comparable to that of the VGG model, but the average inference time, parameter count, and model size are only 15.61%, 33.44%, and 33.40% of the VGG model, respectively. Compared with using only the AVGG model to recognize coal-rock images, using the Bayesian model to fuse drilling geological information improves accuracy by 1.85%, reaching 97.31%.
Research on deformation features and control strategies of repeated mining roadways in Guanjiaya Coal Mine
ZHAO Jie, ZHANG Ningbo, LIU Haibing
2024, 50(8): 44-51. doi: 10.13272/j.issn.1671-251x.2024050043
<Abstract>(63) <HTML> (18) <PDF>(1)
Abstract:
The surrounding rock of the repeated mining roadway is severely deformed and cannot be reused, and the repeated mining roadway has obvious overlapping extension features during the service period. In order to solve the above problems, this study takes the 13092 roadway of Guanjiaya Coal Mine as the research background, and adopts on-site measurement, numerical simulation, and theoretical analysis methods to investigate the overlapping extension features and control measures of repeated mining roadway deformation. The analysis of deformation features of repeated mining roadways shows the following points. ① Under a single mining disturbance, the deformation of repeated mining roadways exhibits zoning and asymmetric failure features, which can be divided into rapid deformation zone, strong deformation zone, and slow deformation zone. The crack damage mainly occurs in the coal wall and coal pillar walls, with less damage to the roof and floor, manifested as significant fragmentation and inward movement of the two sides of the roadway. Severe deformation occurs at the intersection of the coal wall and roof, as well as the coal pillar and floor. ② The secondary mining roadway expands and overlaps on the basis of the primary damage, making the asymmetric damage more significant and forming a butterfly shaped plastic failure zone in the surrounding rock of the roadway. ③ The key time for controlling the surrounding rock of the repeated mining roadway is the first mining stage. The key area is the coal pillar side of the roadway in the strong deformation zone and the slow deformation zone. By analyzing the butterfly deformation features and failure zoning rules of mining roadways, a multi-level coupling control technology for repeated mining roadways is proposed. Shallow low pressure - deep high pressure grouting is used to improve the support force of coal pillars. The anchor cables are used to reinforce and improve the support force of support bodies, achieving coupling control. The comparative analysis of deformation before and after reinforcement has verified that multi-level coupling control meets the requirements of roadway reuse.
Deformation and failure law and control of surrounding rock in the large section chamber of Ulan Mulun Coal Mine
CHEN Ying, YANG Hongtao, SHI Mingzhe, BAO Shiji, ZHANG Zikai, KONG Derui
2024, 50(8): 52-60. doi: 10.13272/j.issn.1671-251x.2024060090
<Abstract>(61) <HTML> (19) <PDF>(2)
Abstract:
In response to the deformation and failure of surrounding rock in large section chamber underground roadways of coal mines, this study focuses on the sorting and filling of large section chamber underground roadways in Ulan Mulun coal mine. Similar simulation experiments are conducted using monotonically increasing and constant load uniaxial compression methods to investigate the deformation and failure laws of surrounding rock in large section chamber underground roadways. The results show the following points. ① The failure evolution and deformation displacement trends of the two loading methods are similar in the compaction stage, elastic deformation stage, and micro fracture stable development stage. ② The sample using monotonically increasing loading method has fewer cracks but larger crack gaps. The sample suddenly breaks along the main crack, during which a large amount of debris flies out. The deformation of the sample is mainly concentrated at the boundary of the surrounding rock. More energy is released during the failure, but the duration of energy release after the peak is relatively short. ③ The stress of the specimen loaded with a constant load remains constant, and the strain slowly increases, during which a large number of small cracks are generated. The deformation position of the specimen mainly surrounds the chamber, and the energy released during failure is relatively small, but the duration of energy release after the peak is longer. Based on the deformation and failure law of the surrounding rock of the large section chamber, a bolt and cable support scheme is proposed. A long anchor cable is installed at the top of the chamber to connect the roof and the hard rock above it as a whole. The inclined anchor rods are installed at the coal rock interface of the roadway to tightly connect the coal rock interface with the surrounding rock mass. The numerical simulation results show that after support, the stress, displacement, and plastic zone of the surrounding rock are significantly reduced, the stability of the surrounding rock is greatly improved, and the support effect is good.
Study on the influence of close range coal seam mining relationship on the deformation law of surrounding rock in lower roadway
ZHANG Xiaojun, SUN Jiarui, MA Yang
2024, 50(8): 61-68. doi: 10.13272/j.issn.1671-251x.2024060079
<Abstract>(73) <HTML> (29) <PDF>(6)
Abstract:
When the mining relationship between the upper coal seam working face and the lower coal seam roadway in the close distance coal seam changes, the deformation and instability mechanism of the surrounding rock of the roadway will be more complicated. At present, there is little research on the dynamic evolution law and instability characteristics of the roadway when the upper coal seam working face and the lower coal seam roadway advance in different directions. Taking the close range coal seam of Nengdong Coal Mine in northern Shaanxi as the research object, a combination of theoretical analysis, numerical simulation, and on-site measurement is used to study the stability of the lower coal seam roadway after the upper coal seam working surfaceis mined. Theoretical analysis shows that the depth of the floor cracks generated after mining the upper coal seam working surfaceis 22.5 m, and they have not developed to the lower coal seam. According to the spatial relationship of mining, the mining face and the roadway are divided into three states: facing, intersecting, and advancing in the opposite direction. Numerical simulations show that when the spatial relationship between the roadway and the working face changes, the deformation of the surrounding rock of the lower coal seam roadway is affected. The results show the following points. ① When the mining relationship between the upper coal seam working surface and the lower coal seam roadway is intersecting and advancing in the opposite direction, the stress of the surrounding rock of the roadway shows a trend of first increasing, then decreasing, and then increasing again. When the length of intervals of travel is 90 m, the maximum stress value is 6.5 MPa, and the stress concentration factor is 1.49. When the length of intervals of travel is 100-110 m, the stress reduction of the surrounding rock of the roadway is the largest, decreasing by 53.2%. When the length of intervals of travel is 150 m, the minimum is 0.95 MPa, and then it continues to increase until it returns to the original rock stress. ② The displacement of the surrounding rock in the roadway increases significantly when the length of intervals of travel is between 100-150 m, and reaches its maximum displacement of 0.036 m at 150 m. As the roadway approaches the boundary coal pillar, the displacement of the roadway decreases. The on-site measurement results show that when the upper coal seam working surface passes through the lower coal seam roadway, the displacement of the roadway increases significantly, and the maximum displacement of the roof is 3.41 cm. It is consistent with the numerical simulation results. If the geological conditions are simple during the process of intersecting advancement, the advancement speed can be appropriately accelerated to reduce the impact of upper coal seam working surface mining on lower level roadways.
Achievements of Scientific Research
Structural performance monitoring of mine hoist head sheave based on digital twins
ZHANG Wenhao, WU Juan, RUAN Kaiyi
2024, 50(8): 69-75. doi: 10.13272/j.issn.1671-251x.2024050086
<Abstract>(92) <HTML> (24) <PDF>(11)
Abstract:
Currently, most of the monitoring research on the head sheave of mine hoists focuses on monitoring the vibration, temperature, and deflection of the head sheave, while there is relatively little research on monitoring the structural performance of the head sheave. In order to solve the above problems, a structural performance monitoring method of mine hoist head sheave based on digital twins is proposed. Based on the actual conditions during the operation of the mine hoist head sheave, a digital twin monitoring system for the mine hoist head sheave is designed. The system consists of an object entity layer, a twin model layer, a twin data layer, an application layer, and connections between each layer. The predicted data in the twin data layer is the real-time head sheave structural performance data predicted by the head sheave structural performance prediction model during the operation of the mine hoist head sheave, including stress and strain data. The structural performance prediction model of the mine hoist head sheave is constructed using a combined surrogate model. A single surrogate model of radial basis function (RBF) is trained using processed finite element data. The weight of the single surrogate model in the combined surrogate model is obtained based on generalized mean square error, thus obtaining the head sheave structural performance prediction model. Taking the five rope friction lifting system of the vertical shaft as the experimental object, based on the Unity3D platform, a digital twin monitoring system for the mine hoist head sheave is established through the construction of virtual space, data transmission, and application modules. The experimental results show that during the operation of the head sheave, the average determination coefficient of the measured strain and predicted strain at the four test points is 0.973 98, indicating a high correlation between the predicted strain and the measured strain. This verifies that the designed prediction model can meet the needs of monitoring the structural performance of the head sheave.
Research on intelligent design of coal mine roadway support scheme
CHEN Wanhui, GUO Rui, HAN Wei, SONG Yongming, LIANG Yanxiang, LIU Yao, WANG Jiaming, XU Na, MENG Bo
2024, 50(8): 76-83, 90. doi: 10.13272/j.issn.1671-251x.2024060044
<Abstract>(61) <HTML> (15) <PDF>(2)
Abstract:
Currently, the design of coal mine roadway support schemes is still mainly based on manual design, engineering analogy, and FLAC model simulation, which has problems such as strong subjectivity, low universality, and insufficient utilization of coal mine support big data. The design method based on expert systems has cumbersome rule setting procedures, large engineering quantities, and low intelligence. Case based reasoning (CBR) and deep learning techniques are introduced into the field of roadway support scheme design. Based on text big data such as coal mine support regulations, support standards, and coal mine roadway geological reports, an intelligent design method for coal mine roadway support scheme is proposed. The method obtains 346 sets of roadway support data from different coal mines, extracts structured data and divides it into input and output parameters, and optimizes the input and output parameters through constant attribute variable filtering and high correlation filtering methods. The method establishes a CBR model and imports the extracted structured data into the CBR model to form a case library of support scheme comparison and selection. The method calculates the similarity between the new roadway support scheme and the historical scheme, and outputs the three historical schemes with the highest similarity for comparison, achieving similar case comparison. BP neural network and long short term memory (LSTM) network are respectively used to establish automatic generation models for coal mine roadway support schemes. By comparing the prediction indicators, it is determined to use the combination of LSTM model and CBR model to establish an intelligent design system for coal mine roadway support scheme. The system is used for the design of auxiliary transportation roadway support scheme in the F6226 working face of Buliangou Coal Mine excavation. Through experiments, it is verified that the deformation of the two sides of the roadway and the maximum displacement of the roof under the system generated scheme are smaller than those under the manual design scheme. The integrity of the roadway roof and two sides is good, the bearing capacity of the surrounding rock is enhanced, and the support effect is significant.
Research on the transportation model and coal quantity calculation algorithm of scraper conveyor based on array
YIN Rui, ZHANG Dongxue, NI Qiang
2024, 50(8): 84-90. doi: 10.13272/j.issn.1671-251x.2024070052
<Abstract>(88) <HTML> (34) <PDF>(6)
Abstract:
Currently, most research on coal quantity detection focuses on the coal quantity detection and recognition of underground belt conveyors in coal mines. The coal quantity detection of scraper conveyors in fully mechanized working (caving) faces only stays at the transfer machine, where infrared scanning devices are installed. The detection technology is single, and because the transfer machine is located at the coal unloading point of the scraper conveyor, the infrared scanning device detects the coal loading of the transfer machine and cannot directly reflect the real-time coal loading on the scraper conveyor, resulting in significant lag. In order to solve the above problems, a transportation model and coal quantity calculation algorithm of scraper conveyor based on array is proposed. This algorithm sets the scraper conveyor as a continuous coal loading carrier, establishes a scraper conveyor transportation model through a continuous array, and characterizes the coal quantity per unit length. Combining the operating speed, drum height, cutting depth and position of the shearer and the operating speed and coal loading factor of the scraper conveyor, the real-time simulation of the unit coal quantity of the scraper conveyor is realized through the method of multi-parameter mathematical modelling. It can intuitively reflect the coal mining process of the underground coal mines and accurately calculate the real-time coal quantity of the scraper conveyor. The results of underground industrial tests show that the algorithm is continuous and reliable, and can accurately calculate the real-time coal quantity on the scraper conveyor. The distribution of coal quantity is close to the ideal state, and it has high convergence and robustness.
Analysis Research
Foreign object detection and counting method for belt conveyor based on improved YOLOv8n+DeepSORT
CHEN Tengjie, LI Yong'an, ZHANG Zhihao, LIN Bin
2024, 50(8): 91-98. doi: 10.13272/j.issn.1671-251x.2024070043
<Abstract>(108) <HTML> (22) <PDF>(12)
Abstract:
The existing foreign object detection methods for belt conveyors have problems such as weak capability to extract object semantic information, poor detection precision, and only recognizing and detecting foreign objects. The methods cannot accurately calculate the number of foreign objects. In order to solve the above problems, a foreign object detection and counting method for belt conveyors based on improved YOLOv8n+DeepSORT has been designed. The method improves the YOLOv8n model and then uses the improved YOLOv8n model to recognize foreign objects in belt conveyors. The method uses the foreign object detection results of the improved YOLOv8n model as input for the DeepSORT algorithm to achieve foreign object tracking and counting on belt conveyors. YOLOv8n improvement method is replacing the C2f module in the backbone network with the C2f_MLCA module to improve the network's information extraction capability in a single color information environment. The method improves the head section using the separated and enhancement attention module (SEAM) to enhance the detection precision of foreign objects when they are obstructed. The method uses Focaler IoU optimization loss function to solve the problem of large differences in the shape of detection objects. The performance verification experiment results of MSF-YOLOv8n model show that the mAP50 of MSF-YOLOv8n model reaches 93.2%, which is 2.1% higher than the basic model. The parameter count is only 2.82×106, which is 0.19×106 less than the basic model, making it more suitable for deployment in edge devices such as inspection robots. The detection precision is 2.2%, 1.3%, and 0.3% higher than YOLOv5s, YOLOv7, and YOLOv8s algorithms, respectively. Although its frame rate is lower than YOLOv8s and YOLOv8n, it still meets the requirements of real-time video detection. The results of foreign object detection and counting experiments show that the DeepSORT algorithm has an accuracy rate of 80% and can accurately track occluded anchor rods and objects with significant shape differences.
Foreign object detection of coal mine underground conveyor belt based on Stair-YOLOv7-tiny
MEI Xiaohu, LYU Xiaoqiang, LEI Meng
2024, 50(8): 99-104, 111. doi: 10.13272/j.issn.1671-251x.18172
<Abstract>(35) <HTML> (17) <PDF>(4)
Abstract:
The existing methods for detecting foreign objects in underground coal mine conveyor belts have poor adaptability to complex scenarios, cannot meet real-time and lightweight requirements, and perform poorly when dealing with foreign objects with large size differences. In order to solve the above problems, a Stair-YOLOv7-tiny model is proposed based on the lightweight YOLOv7-tiny model for improvement, and applied to the detection of foreign objects in coal mine underground conveyor belts. This model adds feature concatenation units to the efficient layer aggregation network (ELAN) module to form a Stair-ELAN module. The model fuses low dimensional features from different levels with high-dimensional features, strengthens the direct connection between feature levels, enhances information capture capabilities, and strengthens the model's adaptability to objects of different scales and complex scenes. The introduction of Stair-head feature fusion (Stair-fusion) for detection heads forms a Stair-head module. The model enhances the feature expression capability of medium and low resolution detection heads by fusing detection head features of different resolutions layer by layer, achieving complementary feature information. The experimental results show that the Stair-YOLOv7 tiny model has better detection performance than CBAM-YOLOv5, YOLOv7 tiny, and its lightweight model on the open-source dataset CUMT BelT for conveyor belt foreign objects. The accuracy, average precision, recall, and precision are 98.5%, 81.0%, 82.2%, and 88.4%, respectively, and the detection speed is 192.3 frames per second. In the video analysis of conveyor belt monitoring in a certain mine, the Stair-YOLOv7-tiny model does not have any missed or false detection, achieving accurate detection of foreign objects in the conveyor belt.
Small object detection method for mining face based on improved YOLOv8n
XUE Xiaoyong, HE Xinyu, YAO Chaoxiu, JIANG Ze, PAN Hongguang
2024, 50(8): 105-111. doi: 10.13272/j.issn.1671-251x.2024060013
<Abstract>(62) <HTML> (32) <PDF>(4)
Abstract:
In order to effectively detect and recognize whether the personnel on the mining face in coal mines are wearing safety protection devices, a small object detection method based on improved YOLOv8n is proposed. It is applied in situations such as poor underground lighting conditions, small object sizes of safety protection device, and similar colors to the background. The method integrates Dynamic Snake Convolution (DSConv) into the C2f module of YOLOv8n backbone network to construct a C2f DSConv module, in order to enhance the model's capability to extract multi-scale features. The method introduces polarized self-attention (PSA) mechanism in the Neck layer to reduce information loss and improve feature expression capability. The method adds one detection head specifically designed for small objects at the Head layer, forming a four detection head structure to expand the detection range of the model. The experimental results show that the improved YOLOv8n model has an average precision of 98.3%, 95.8%, 89.9%, 87.2%, and 90.8% for detecting underground personnel and their safety helmets, mining lights, masks, and self rescue devices, respectively. The average precision is 92.4%, which is better than Faster R-CNN, YOLOv5s, YOLOv7, and YOLOv8n models. The detection speed reaches 208 frames per second, meeting the requirements of object detection precision and real-time performance in coal mines.
A coal mine underground drill pipes counting method based on improved YOLOv8n
JIANG Yuanyuan, LIU Songbo
2024, 50(8): 112-119. doi: 10.13272/j.issn.1671-251x.2024040073
<Abstract>(112) <HTML> (32) <PDF>(6)
Abstract:
In order to improve the efficiency and precision of underground drill pipe counting in coal mines, a coal mine underground drill pipe counting method based on the improved YOLOv8n model is proposed. The YOLOv8n-TbiD is established.The model can accurately detects and segments drill pipes in mine drilling rig working videos. The main improvements include the following points. In order to effectively capture the boundary information of drill rods and improve the precision of the model in recognizing drill rod shapes, the weighted bidirectional feature pyramid network (BiFPN) is used instead of the path aggregation network (PANet). To address the issue of drill pipe objects being easily confused with dim mine environments, Triplet Attention is added to the SPPF module of the Backbone network to enhance the model's capability to suppress background interference. In response to the small proportion of drill pipes in the image and the complexity of background information, the Dice loss function is used to replace CIoU loss function to optimize the segmentation processing of drill pipe objects in the model. The method uses the YOLOv8n-TBiD model to segment the drill pipe and its mask information. A drill pipe counting algorithm is designed based on the rule that the mask area of the drill pipe decreases during drilling and suddenly increases when a new drill pipe is installed. The working video of the drilling rig in the fully mechanized working face is selected, in order to conduct experimental verification of drill pipes counting method based on YOLOv8n-TBiD model. The experimental results show that the mean average precision of the YOLOv8n-TBiD model for detecting drill pipes reaches 94.9%. Compared with the comparative experimental models GCI-YOLOv4, ECO-HC, P-MobileNetV2, YOLOv5, and YOLOX, the accuracy increases by 4.3%, 7.5%, 2.1%, 6.3%, and 5.8%, respectively, and the detection speed increases by 17.8% compared to the original YOLOv8n model. The proposed drill pipe counting algorithm achieves precision of 99.3% on video datasets from different underground coal mine environments.
Coal and gangue segmentation and recognition method based on YOLOv5-SEDC model
YANG Yang, LI Haixiong, HU Miaolong, GUO Xiucai, ZHANG Huipeng
2024, 50(8): 120-126. doi: 10.13272/j.issn.1671-251x.2024010078
<Abstract>(73) <HTML> (26) <PDF>(7)
Abstract:
The existing coal and gangue segmentation and recognition technology has a large number of parameters, slow classification speed, and low recognition accuracy. The YOLOv5-seg model is prone to losing texture details and grayscale feature information on the image surface during up and down sampling operations, which reduces the efficiency of coal and gangue recognition. The YOLOv5-seg model overly focuses on global features during training, while neglecting the locally significant regions and features that are crucial for coal and gangue recognition. In order to solve the above problems, a coal and gangue segmentation and recognition method based on YOLOv5-SEDC model is proposed. Firstly, the method receives an image containing the shape information of coal and gangue, and uses the backbone network for feature extraction to generate a feature map. The method integrates the SENet module into the YOLOv5-seg model to preserve the texture details and grayscale features of coal and gangue surfaces, avoiding information loss caused by down sampling. The method adopts a dilated convolution strategy with different dilation rates instead of traditional convolution kernels. It not only expands the receptive field of the model, but also effectively reduces the number of model parameters. Finally, the segmentation detection head finely processes the fused features to achieve precise segmentation and recognition of coal and gangue. A coal and gangue image acquisition experimental platform is established at the actual coal and gangue sorting site of Daliuta Coal Mine. The ablation experiment results show that the accuracy of coal and gangue recognition of YOLOv5-SEDC model is improved by an average of 1.3% compared to YOLOv5-seg model. The parameter quantity is reduced by 0.7×106, and the detection speed is increased by 1.4 frames/s. The comparative experimental results show the following points. ① The accuracy of the YOLOv5-SEDC model is improved by 10.7%, 2.7%, 1.9% compared to the YOLOv3-tiny, YOLOv5-seg, and Mask-RCNN models, respectively, reaching 95.8%. ② The recall rate of the YOLOv5-SEDC model has increased by 3.0%, 2.1%, and 0.9% compared to the YOLOv3-tiny, YOLOv5-seg, and Mask-RCNN models, respectively, reaching 89.1%. ③ The mAP of the YOLOv5-SEDC model has increased by 6.4%, 6.3%, and 1.8% compared to the YOLOv3-tiny, YOLOv5-seg, and Mask-RCNN models, respectively, reaching 95.5%. ④ The F1 value of the YOLOv5-SEDC model has increased by 5.2%, 4.2%, 2.1% compared to the YOLOv3-tiny, YOLOv5-seg, and Mask-RCNN models, respectively, reaching 92.2%. ⑤ The detection speed of the YOLOv5-SEDC model is reduced by 1.9, 1.4, and 2.7 frames/s compared to the YOLOv3-tiny, YOLOv5-seg, and Mask-RCNN models, respectively. The visualization results show that the YOLOv5-SEDC model has higher detection accuracy for coal and gangue than the YOLOv5-seg and Mask-RCNN models. It indicates that the YOLOv5-SEDC model has good performance in coal gangue segmentation and recognition.
Coal rock crack recognition method based on connectivity threshold segmentation
XIAO Fukun, LIU Huanhuan, SHAN Lei
2024, 50(8): 127-134. doi: 10.13272/j.issn.1671-251x.2024050092
<Abstract>(90) <HTML> (27) <PDF>(8)
Abstract:
The development morphology of coal rock cracks is an important factor affecting the permeability of coal rock and determining the mechanical features of coal rock mass. A coal rock crack recognition method based on connectivity threshold segmentation is proposed to address issues such as improper handling of complex structures, insufficient preservation of crack boundary features, and noise interference in the process of recognizing coal rock cracks. Firstly, histogram equalization enhancement algorithm and non local mean filtering denoising algorithm are used to preprocess the image. Secondly, adaptive Otsu threshold segmentation is used to determine the threshold of the preprocessed image, recognize possible crack areas, and apply morphological operations to refine these areas, further highlighting the boundary features of cracks. Thirdly, seed points are extracted by Canny edge computing to recognize key features in the image. Finally, based on these seed points, regional growth operations are performed to effectively suppress noise and highlight crack information more clearly while smoothing image cracks. The experimental results show the following points. ① The mean square error of connectivity threshold segmentation is reduced by an average of 7.20 and 7.10 dB compared to adaptive Otsu threshold segmentation and adaptive threshold segmentation, respectively. The peak signal-to-noise ratio of connectivity threshold segmentation is improved by an average of 0.60 and 0.59 dB compared to adaptive Otsu threshold segmentation and adaptive threshold segmentation, respectively. ② Connectivity threshold segmentation not only effectively solves the problems of unclear crack extraction, poor end extraction performance, and disappearance of connection features, but also significantly reduces the interference of noise, making crack features more prominent, thereby greatly improving the accuracy and completeness of crack recognition. ③ On the basis of adaptive Otsu threshold segmentation, connectivity threshold segmentation enhances crack features and effectively eliminates noise points. The average accuracy is improved by 8% and 0.8% respectively compared to adaptive threshold segmentation algorithm and adaptive Otsu threshold segmentation, reaching 98.9%.
Research on pitch control of coal mine roadheader based on fuzzy neural network PID
MAO Qinghua, CHEN Yanzhang, MA Cheng, WANG Chuanwei, ZHANG Fei, CHAI Jianquan
2024, 50(8): 135-143. doi: 10.13272/j.issn.1671-251x.2024070014
<Abstract>(67) <HTML> (23) <PDF>(7)
Abstract:
Currently, PID control method is mainly used for the pitch control of coal mine roadheader, and the control precision is not high in the case of time-varying and nonlinear hydraulic system during the pitch control of roadheader. The pitch control of roadbeader is realized by controlling the stroke of the hydraulic cylinder. Combining the traditional PID algorithm with fuzzy control and neural network, the accuracy of the stroke control of the hydraulic cylinder can be effectively improved. In order to solve the above problems, a pitch control method for coal mine roadheader based on fuzzy neural network PID is proposed. By analyzing the kinematic relationship of the support part of the roadheader, the mathematical relationship between the pitch angle and the hydraulic cylinder of the support part is obtained. The working principle of the pitch control hydraulic system of the roadheader is introduced, and the hydraulic system and its transfer function model are established. The method combines fuzzy control with neural networks to form a fuzzy neural network. The method optimizes PID control parameters by using the fuzzy neural network. The method combines the mathematical model of the support mechanism and the transfer function model of the hydraulic system to establish a fuzzy neural network PID control model for the pitch angle of the roadheader. It achieves automatic and precise control of the pitch mechanism of the coal mine roadheader. This method can make the pitch mechanism of the roadheader reach the preset position more quickly and accurately, solving the time-varying and nonlinear problems in the pitch control of roadheader. The simulation results show that the fuzzy neural network PID control algorithm reduces tracking errors by 69.34% and 74.49% respectively compared to fuzzy PID and PID control algorithms. The method simulates the pitch control of coal mine roadheaders under sudden and following working conditions through hydraulic cylinder displacement control. The results show that compared with fuzzy PID and PID control algorithms, the fuzzy neural network PID control algorithm has the smallest pitch control tracking error, shortens the average response time to position signals by 27.22% and 50.33% respectively, and has better dynamic control performance.
Channel estimation method for IRS assisted mine communication system based on self supervised learning
WANG Anyi, LI Xinyu, LI Mingzhu, LI Ruoman
2024, 50(8): 144-150. doi: 10.13272/j.issn.1671-251x.2024070038
<Abstract>(61) <HTML> (18) <PDF>(2)
Abstract:
A channel estimation method for intelligence reflecting surface (IRS) assisted mine communication system based on self supervised learning (SSL) is proposed to address the problems of multipath fading, non line of sight communication, and difficulty in obtaining true labels caused by complex mine environments. The method builds an underground communication system model based on the Nakagami-g fading channel model and IRS signal transmission model, and solves the problems of multipath fading and non line of sight communication through IRS technology. Preliminary channel estimation is performed using the least squares (LS) algorithm, and then the channel estimation results are optimized using octave convolution (OCT) neural network under the SSL framework. OCT directly processes both high-frequency and low-frequency components, capturing both the rough features and subtle differences of the channel, providing comprehensive channel information, and thus more accurately estimating the channel state. The SSL algorithm uses received signals and their noisy versions as training data to improve the precision and efficiency of IRS assisted channel estimation through the intrinsic structure of unlabeled data, thereby reducing reliance on manual labeling. The simulation results show the following points. ① Introducing IRS technology can effectively reduce channel estimation errors. ② The loss value of OCT neural network is significantly lower than that of CNN, and the data fitting effect is better. OCT neural network has high computational efficiency and can improve the overall performance of channel estimation in communication systems. In environments with limited computing resources, OCT neural networks can maintain low parameter and memory usage. ③ The SSL algorithm can maintain a low normalized mean square error under all signal-to-noise ratio conditions, verifying its efficiency and robustness in channel estimation. ④ The channel estimation method for IRS assisted mine communication system based on SSL has good scalability and robustness in large-scale networks.
Dynamic feature extraction for flotation froth based on centroid-convex hull-adaptive clustering
WEI Kai, WANG Ranfeng, WANG Jun, HAN Jie, ZHANG Qian
2024, 50(8): 151-160. doi: 10.13272/j.issn.1671-251x.18182
<Abstract>(40) <HTML> (13) <PDF>(2)
Abstract:
In the face of complex flotation site environments and issues such as unclear boundaries caused by the mutual adhesion of flotation froth, existing methods for extracting dynamic features (such as flow velocity and collapse rate) often fail to accurately delineate the dynamic feature sampling regions corresponding to each froth, cannot comprehensively match feature points between adjacent frames, and have difficulty effectively identifying collapse regions. To address these problems, a dynamic feature extraction method for flotation froth based on a centroid-convex hull-adaptive clustering approach is proposed. This method employs an improved Mask2Former, integrated with the multi-scale feature extraction capability of Swin-Transformer, to accurately locate froth centroids and effectively identify collapse regions. An optimal convex hull evaluation function is used to search for the convex hull formed by the centroids of adjacent froth surrounding the target froth, thereby fitting a dynamic feature sampling region close to the actual froth contour. The local feature matching with transformer (LoFTR) algorithm is applied to match feature point pairs between adjacent frames. For all feature point pairs within the dynamic feature sampling region, the main flow velocity of each froth is extracted using the main feature adaptive clustering method based on the OPTICS algorithm. Experimental results show that this method achieves accuracy rates of 88.83% and 97.92% and intersection over union (IoU) rates of 77.90% and 96.52% in ordinary froth centroid location and collapse region identification tasks, respectively. It also achieves a correct feature point pair matching rate of 99.93% with an average exclusion rate of 2.69%. The method effectively delineates feature sampling regions close to the actual froth boundaries under various conditions, enabling the quantitative extraction of each froth's dynamic features.

苏ICP备 05016349号-2

Supported by: Beijing Renhe Information Technology Co. Ltd.  Visits：