Large quantities of coal gasification slag are generated during the exploitation and utilization of coal resources, and the efficient and green disposal of this slag is one of the key issues that urgently needs to be resolved in the coal sector. Through an orthogonal experimental system, the working and mechanical properties of coal gasification slag paste filling materials were systematically studied, and their compliance with current standard specifications was compared. Using range analysis, the optimal mix proportion was determined as follows: a slag-to-gypsum ratio of 2.5∶1.5, a fly ash blending ratio of 2∶2, a slurry mass concentration of 76%, and a corresponding coal gasification slag content of 11.7%. The static leaching method was adopted to conduct immersion tests on this material, and the water quality indicators of the leachate at different immersion ages were examined. The results show that the indicators of the leachate stabilize in the later stage, comply with the Class III groundwater quality standard, and meet the basic requirements for groundwater quality. The F-value method was employed for a comprehensive water quality assessment. The results indicate that the groundwater quality grade for different immersion durations is predominantly "Good". Although the water quality declines slightly with prolonged immersion time, it remains in a safe state overall. Compared with traditional paste filling materials, the cost of the coal gasification slag paste filling material is reduced by 16%, and the comprehensive performance is significantly enhanced. This provides an important basis for the engineering application of coal gasification slag paste filling technology and holds great significance for realizing the resource utilization of coal-based solid waste and the protection of groundwater quality.

In view of the problems of large water inflow and difficult grouting and water plugging in some underground mines in the copper (cobalt) metallogenic belt of Central Africa, which seriously restrict the safe mining of mineral resources, Chambishi Copper Mine was selected as the research object, and efficient plugging materials and processes suitable for this mine and similar water-rich mines were developed. Five kinds of materials, including local clay, bentonite, a chemical material, domestic ultrafine cement and tailings from Africa, were selected for laboratory tests. Multi-proportion slurries were prepared according to different mixing schemes, and the physicochemical properties (viscosity, relative density, bleeding rate and plastic strength), setting time and compressive strength of each slurry's stone body were tested. The results show that the local mud cake meets the requirements for pulping. The 7 d compressive strength of the modified clay slurry stone body is 0.11-0.69 MPa, and its viscosity increases significantly with a decrease in water-solid ratio. It is suitable for large-scale ground curtain, wellbore curtain and other projects, and can reduce material costs by about 30%. The local bentonite meets the requirements for pulping. The 7 d compressive strength of the cement-bentonite slurry stone body is 0.10-2.34 MPa. When the water-solid ratio is ≤2∶1, the setting time shortens with the increase in bentonite content, making it suitable for deep poorly injectable strata. In the deep multi-layer sand belt area, its injectability is 7 times higher than that of traditional cement slurry. The 28 d compressive strength of the ultrafine cement slurry stone body can reach up to 59.8 MPa, and its strength decreases with the increase in water-binder ratio. It is suitable for micro-fracture and backfill grouting, offering improved impermeability and high compressive strength. By adjusting the content of GROUTEX, the chemical slurry can achieve instant setting, and the compressive strength of its stone body reaches 43.65 MPa. It is suitable for sealing poorly sealed boreholes, achieving instant setting with a strength greater than 43 MPa, thereby preventing water inrush accidents caused by exposing water-rich boreholes underground. The bleeding rate of the tailings slurry is high, and the 7 d compressive strength of its stone body reaches up to 9.7 MPa; further improvement tests with different additives are therefore necessary.

To solve the problem that the cultural relic area and cultural relic group directly above a copper ore body are affected by mining blasting vibration, the dynamic response and parameter optimization of blasting in near-surface mining were systematically studied. Firstly, the relationship between the surface cultural relic area and the ore body was analyzed, and the maximum allowable blasting vibration velocity was determined by combining domestic and international safety criteria for blasting vibration. Based on the M. A. Sadovsky empirical formula, the maximum single-delay blasting explosive charge within the blast safety distance was obtained. Then, blasting simulations of the four levels at 1 945 m, 2 020 m, 2 095 m and 2 245 m closest to the surface throughout the mine's life cycle were carried out by numerical simulation, and the rationality of the maximum single-delay blasting explosive charge was verified. Finally, during blasting at the representative level underground, surface vibration tests were conducted using measuring points arranged on the surface. The results show that the maximum single-delay explosive charges at the four level are 745 kg, 390 kg, 172.5 kg and 10 kg, respectively, all of which are equal to or less than the allowable maximum single-delay explosive charge. The surface peak vibration induced by blasting velocity at the 2 245 m level is 0.5 cm/s, which is greater than that of the other three levels, but remains less than the maximum allowable vibration velocity, indicating that the blasting seismic wave has no effect on the surface structures. Based on the field blasting vibration data, an SSA-Elman neural network model for predicting Y-direction blasting vibration was constructed and compared with the traditional Elman neural network model. The predicted values of the SSA-Elman neural network model are basically consistent with the measured values, with an average MAE of 0.01, which verifies its accuracy. The research results provide a basis for the design of blasting parameters for other levels of the copper mine.

To satisfy the current demand for the construction of information systems supporting the dual prevention mechanism—specifically, hierarchical risk control and hidden danger investigation and governance—in non-coal mines, a Browser/Server (B/S) architecture-based information system tailored for non-coal underground mines was developed. In accordance with the implementation workflow of the dual prevention mechanism, the system was grounded in the "9+X risk identification method" and the "human-centric evaluation method", and adopted a business architecture comprising "one platform, two terminals, and three levels". Through data collaboration between the platform and terminals, it had realized entire-process refined and intelligent construction and management of the dual prevention mechanism in mining enterprises, which effectively promoted transparent risk control and flattened hidden danger governance. The application results in pilot mines demonstrate that the completion rate of hidden danger investigation and governance reaches 100%, and management efficiency increases by 90%. The system significantly enhances mine safety management efficacy, facilitates the implementation of the all-staff safety responsibility system, drives the digital transformation of safety management, and provides robust support for the construction of work safety standardization in non-coal mining enterprises.

In view of the threat of complex karst water disasters in deep mining of metal mines, it is urgent to develop refined curtain closure water control technology and multi-dimensional effect evaluation methods. In order to improve the efficiency of water disaster prevention and control and verify the engineering effect, ZK Iron Mine was selected as the research object. Combined with geophysical prospecting and drilling to identify the karst channel in advance, the fine grouting technology based on channel identification was implemented (dynamic control pressure of 2-4 MPa, water-cement ratio of 1∶1-0.6∶1). FEFLOW software was used to simulate and predict the change of groundwater flow field before and after curtain closure and the attenuation effect of daily mine drainage (reduced to 7 984 m³/d, water plugging efficiency of 65%). A multi-dimensional effect evaluation system integrating grouting volume analysis, hydrological monitoring, geophysical exploration, inspection hole verification and AHP quantitative evaluation was constructed. The results show that the fine grouting process effectively plugs the complex karst fracture network. The daily average drainage volume of the mining area decreased from 22 984 m³/d to 7 770 m³/d, a decrease of 15 214 m³/d, and the water plugging efficiency reached 66.2%, which is close to the simulated predicted value. The comprehensive evaluation score of AHP is 0.895 (excellent grade), among which the reduction of drainage rate (weight of 31.1%) and the contribution of water plugging efficiency are significant. This technical system provides theoretical support and a replicable technical paradigm for aquifer protection and water disaster prevention and control in similar metal mines.

During the mining process of the mine working face, the disturbance of the aquifer can easily lead to an increase in the water inflow of the working face, which threatens the safe production of the mine. Accurate prediction of the water inflow of the working face can effectively prevent water disaster accidents. In this study, the multi-feature prediction method of water inflow in working face based on deep learning model was studied. Combined with the characteristics of water inflow data and the actual situation of the site, the "double aquifer water level+maximum height of microseismic event" was selected as the characteristic data, and the multi-feature prediction model of water inflow in working face of mine based on VMD-iCHOA-GRU was established. At the same time, the prediction effect of multi-feature model was further verified under different feature combinations, single feature conditions and different working face data. The results show that the evaluation indexes MAE, RMSE and MAPE of the water inflow prediction results of the VMD-iCHOA-GRU model are 53.56 m³/d, 62.98 m³/d and 3.1%, respectively, which are better than other control models, and its prediction accuracy is the highest. The prediction accuracy of the multi-feature model is higher than that of the single-factor model and other single-feature and double-feature models in various combinations, and the stronger the correlation between the feature data and the research target, the more conducive it is to improving the prediction accuracy. The prediction effect of the model is good under different working face data, which verifies that the model has relative stability and certain scalability. The research results provide new ideas and new methods for the prediction of water inflow in mine working faces, and have scientific guiding significance and reference value for mine water prevention and control.

With the expansion of open-pit mining scale, slope stability problems have become increasingly prominent. Great significance is attached to realizing safe and efficient open-pit mining by accurately predicting slope stability and taking preventive measures in advance. A whale optimization algorithm-convolutional neural network-long short-term memory (WOA-CNN-LSTM) prediction model for slope stability was constructed based on factor analysis. Six factors, including cohesion, internal friction angle, slope angle, slope height, pore pressure ratio and unit weight, were selected as the discrimination indicators of the model. Factor analysis was adopted to eliminate the correlation among the six indicators, reduce data dimensions and extract common factors. The extracted common factors were taken as the inputs of the WOA-CNN-LSTM model, and the model was applied to slope engineering cases to verify its reliability. The research results show that the CNN-LSTM network can be optimized by iteratively updating and searching for optimal network parameters using the WOA algorithm. The problems of the CNN-LSTM model, including falling into local optimal solutions and thegradient descent affectingthe prediction accuracy of slope stability, can be effectively overcome. The prediction accuracy and generalization ability of the model are thus improved, and a new idea for slope stability prediction is provided.

To ensure the safe operation of open-pit mines, safety evaluation of their slopes needs to be carried out. A slope of a certain mine was taken as the research object. A numerical calculation model of the slope was established by the finite element method, the distribution laws of slope stress, displacement and plastic zone were analyzed, and the slope safety factor was calculated by multiple quantitative analysis methods. The results show that the maximum slope displacement, plastic zone and minimum safety factor are all close to the third-stage slope. The third-stage slope was taken as the evaluation object. Single index measurement functions were constructed by the unascertained measurement theory, index weights were determined by the entropy weight method, and the distance discriminant method was adopted as the evaluation criterion for slope safety grade. The research results show that the evaluation grade of the slope in an open-pit mine is Grade Ⅱ. The evaluation results are consistent with the numerical simulation results, which verifies that the mine slope is in a relatively stable safety state. After reinforcement and monitoring were carried out on the slope risk areas, no slope accidents occurred during the operation of the open-pit mine, which verifies the reliability and practicability of the unascertained measurement theory-finite element method for slope safety evaluation.

The southeast orebody of Chambishi Copper Mine is hosted in shallow metamorphic rock strata. The surrounding rock features well-developed joints and fissures. Under the influence of high in-situ stress and long-term mining disturbances, the roadways are prone to disasters such as roof fracturing, anchor bolt loosening, and anchor mesh collapse. To address the issues of poor support performance and insufficient adaptability of the existing split-set anchor bolts, an analysis of joint fissures and rock mechanical properties was conducted for the fractured quartzite in the hanging wall and footwall of the orebody. Based on the fracturing and deformation characteristics of typical roadways, the applicability and mechanical properties of different anchor bolt types were comprehensively considered, and resin anchor bolts were proposed to replace the original split-set anchor bolts for supporting roadways in fractured surrounding rock. The 3DEC numerical simulation method was employed to establish a typical roadway model, and the displacement control effects and surrounding rock response characteristics of the two support schemes were compared and analyzed. The results show that the quality grade of the quartzite rock mass is determined to be Class III based on the combined application of the RMR method, Q-system method, and BQ method. When resin anchor bolts with a length of 2.4 m and a spacing and row distance of 1.2 m are used to replace the original split-set anchor bolts for support, the roof subsidence decreases by 8.4%, the floor heave decreases by 4.4%, the inward convergence deformation of the sidewalls is significantly improved, and the overall stability of the roadway is enhanced compared with that achieved by split-set anchor bolts. The industrial application results show that the average pullout force of the resin anchor bolts is 174.17 kN, which is far higher than the design value of 101.47 kN. Although the initial support cost is 2.75 times that of the split-set anchor bolt support, the supplementary support rate is reduced to 2.46%, and the supplementary support cost is reduced by approximately 84%. The research results can provide technical support and a theoretical basis for support optimization in similar roadway engineering.

The tunneling working face is an area prone to frequent gas accidents during coal mining. In view of the strong nonlinear and spatiotemporal dynamic correlation characteristics of gas concentration variations, a prediction model for gas concentration at the tunneling working face, named KAN-iTransformer, was proposed, which integrates the Kolmogorov-Arnold Networks (KAN) and an improved iTransformer (Inverted Transformer). Firstly, a relevant feature dataset for gas concentration prediction was constructed through data preprocessing and Spearman correlation coefficient analysis. Secondly, based on the inverted Transformer structure, the complete variable sequence was regarded as a token, and a graph attention mechanism was introduced to explicitly model the spatial topological relationships of the sensor network, enabling effective modeling of the dynamic dependencies among sensors. Thirdly, the learnable activation functions of KAN were used to replace the traditional linear weights, enhancing the model's ability to characterize complex nonlinear relationships among multi-source monitoring parameters. Finally, a hierarchical response mechanism combining point prediction and interval estimation was proposed to enable advanced dynamic hierarchical response to gas overrun risk. The field test results show that the MSE, MAE and MAPE of the improved KAN-iTransformer model in single-step prediction are 0.000 307, 0.012 921 and 2.321 373%, respectively, significantly outperforming both the iTransformer and Transformer models. In multi-step prediction, the MSE of the improved KAN-iTransformer model is 0.000 913, which is 63.7% and 80.9% lower than those of the comparison models, respectively, demonstrating better long-term prediction performance and effectively realizing accurate prediction and dynamic early warning of gas concentration at the tunneling working face.