In 1956, in accordance with the "Twelve-Year Plan for Scientific and Technological Development" formulated by the former State Science and Technology Commission, preparatory work was conducted for the establishment of the Nonferrous Metal Mining Research Institute, and the institute was founded. In 1962, it was renamed Changsha Institute of Mining Research. It was formerly affiliated with the Ministry of Metallurgical Industry and the China National Nonferrous Metals Industry Corporation, and is now an important science and technology-based enterprise directly managed by China Minmetals Corporation. Over the past 70 years, Changsha Institute of Mining Research has been devoted to integrated comprehensive services covering design and consulting, technology development, inspection, testing and certification as well as engineering construction and operation and maintenance, for the entire process and full life cycle of the development and utilization of mineral resources. It was once ranked 16th in comprehensive strength among more than 800 research institutes administered by former national ministries and commissions, and it is widely recognized as "China's No.1 Institute in Metal Mining". In line with the national strategy for ensuring the supply and security of mineral resources, 1 088 major scientific and technological achievements have been made by Changsha Institute of Mining Research in fields such as mining technology, mine safety, mine backfilling, water prevention and control, digital and intelligent mines, intelligent equipment, testing and inspection, and mineral processing technology. Among them, 890 scientific and technological achievements have been awarded, including 63 National Science and Technology Awards (2 Special Prizes and 4 First Prizes of the National Science and Technology Progress Award). In addition, 1 217 patents have been granted, including 434 invention patents, and 55 national and industry standards have been compiled with the institute as the lead or participating unit. And the related scientific and technological achievements have been promoted and applied to more than 3 000 enterprises at home and abroad. The 70-year scientific and technological development history of Changsha Institute of Mining Research was systematically reviewed, its major scientific and technological achievements and industry contributions were summarized, the practical challenges facing the development of the industry were analyzed, and an outlook on future directions of scientific and technological innovation in the mining field was provided. A reference is thereby provided for the high-quality development and scientific and technological self-reliance and self-strengthening of China's mining industry.

As shallow metal mineral resources become increasingly depleted, advancing into the deep subsurface has become an inevitable strategic choice for safeguarding national resource security. However, the complex geomechanical environment characterized by "three highs and one disturbance" in deep mining poses severe challenges to conventional stoping methods in terms of safety, cost control, and operational efficiency. By reviewing typical engineering practices of deep mines at home and abroad and recent research achievements, the development status, key engineering technical difficulties, and main technological paths of deep metal mining were systematically summarized. The applicability and limitations of deep mining methods under different ore body occurrence conditions was analyzed, and it was pointed out that "large-scale mining, intelligent equipment, and continuous operation" constitute the core direction of deep mining technological transformation. For three typical types of ore body occurrence conditions, namely thick and stable ore bodies, fractured ore bodies, and thin veins in deep mining, technical systems were respectively proposed with high-stress coupled rock fragmentation and modular continuous stoping, environmental reconstruction and stress-relief mining, as well as precise blasting and waste-free efficient stoping as the core. Meanwhile, the application potential of non-explosive mechanized continuous mining technology in reducing disturbance and enhancing inherent safety was also analyzed. This study aims to provide a theoretical basis and technical pathways for the transformation of deep metal mining in China from traditional modes to intelligent and continuous technology systems, thereby contributing to the formation of new quality productive forces in metal mineral resource development.

Deep orebody mining generally faces a complex occurrence environment characterized by high in-situ stress, high ground temperature, high osmotic pressure, and strong mining disturbance. Conventional backfill mining technologies and equipment struggle to meet the demands of safe and efficient mining. Under large-scale deep mining conditions, deep shafe backfill faces problems such as high hydration heat release of cementitious materials, slow backfill formation, insufficient backfill system capacity, and high-potential-energy slurry transport that easily induces partially filled pipe flow, water hammer, and pipe bursting risks. These issues have become key bottlenecks restricting the safe and efficient mining of deep mines. Focusing on the engineering requirements for large-scale, safe, and efficient backfill in deep metal mines, three key issues were proposed: the rapid and sustained low-heat solidification mechanism of ultrafine tailings backfill slurry, the pressure regulation and high-efficiency energy utilization mechanism of full-pipe transport in deep shafts, and the stability control and formation mechanism of high backfill under deep high in-situ stress. Four key technologies were systematically summarized: performance regulation of low-heat, early-strength, and high-strength backfill cementitious materials; large-scale continuous and steady-state thickening of ultrafine tailings; pressure-reducing power generation and full-pipe transport safety assurance for deep mine backfill slurry; and formation and quality assurance of high backfill under deep high in-situ stress. The research and development directions for new materials, technologies, and equipment for green and efficient backfill in metal mines at depths of over 1 000 m and with capacities of tens of millions of tons were clarified, and a large-scale, safe, and efficient backfill technology system for deep mining was constructed. This aims to comprehensively enhance China's capacity for green and efficient development of deep strategic mineral resources and provide a theoretical basis and technical support for the large-scale development of deep mineral resources.

As metal mining extends to depths greater than 1 000 m, the water hazard occurrence environment and disaster-causing process exhibit characteristics such as high confined pressure coexisting with weak water yield, high stress-high water pressure coupling, diverse water sources, concealed water-conducting channels, and significant time-dependent disaster evolution. The traditional prevention and control mode, which is dominated by static hydrogeological condition analysis and passive drainage, can no longer meet the requirements of safe deep mining. Based on the long-term research and engineering practice achievements of Changsha Institute of Mining Research Co., Ltd in mine water hazard prevention and control, this paper systematically summarized four aspects: water hazard characteristics, disaster-causing mechanisms, prevention and control technologies, and coordinated groundwater safety management. Firstly, the types and basic characteristics of water hazards in deep metal mines were sorted out from the dimensions of water sources, water-conducting channels, and disaster-causing modes. Secondly, the mechanisms of initiation, triggering, and evolution of deep water hazards were elucidated from three levels: geological structure-controlled water, rock mass damage-seepage coupling under high stress-high water pressure, and thermal-hydraulic-mechanical-chemical multi-field coupling. Finally, key technologies such as fine hydrogeological exploration, water inrush risk assessment, curtain grouting, sealing of strong dynamic water channels, controlled dewatering, flooded shaft treatment, and multi-source information monitoring and early warning were summarized. A closed-loop prevention and control system of "fine exploration-risk assessment-engineering treatment-monitoring and early warning" was constructed. The research can provide a theoretical basis and engineering reference for the collaborative development of deep metal mine water hazard prevention and control from single disaster treatment to safe mining, groundwater resource utilization, and ecological environment protection.

With the increasing complexity of orebody occurrence conditions and the continuous improvement of resource recovery requirements, complex fractured orebodies have become important targets for safe, efficient, and low-dilution-loss mining in metal and non-metal mines. The backfill mining method provides an important technical approach for the mining of such orebodies by controlling stope exposure, reconstructing the surrounding rock-pillar-backfill bearing relationship, and regulating the mining-backfill cycle. Based on a review of research and engineering applications of backfill mining at home and abroad, starting from the basic conditions and dominant control mechanisms of orebodies, the backfill mining methods for complex fractured orebodies were categorized into the small-exposure continuous mining category, structural unit delayed backfill category, and strong-control combined support backfill category. Furthermore, combined with typical cases, their applicable conditions, control mechanisms, and engineering effects were compared. The research results show that the small-exposure continuous mining category is suitable for the fine mining of thin and narrow orebodies, and orebodies with complex boundaries; the structural unit delayed backfill category can balance the production capacity and stope bearing conversion of medium-thick to massive orebodies; the strong-control combined support backfill category can improve the stability control capability for weak roofs, large spans, and local fractured zones. At the present stage, it is still necessary to focus on solving issues such as the matching of engineering geological conditions and method parameters, the sequential connection of mining and backfilling, the process control of dilution and loss, the composite bearing mechanism, and technical economy. In the future, geological zoning, function-oriented backfill design, collaborative organization of mining-support-backfill-transport, and dynamic optimization driven by monitoring feedback should be strengthened.

As the core technology of green mining in metal mines, unclassified tailings paste backfill technology plays an irreplaceable role in the coordinated development of mineral resource exploitation and ecological environment protection. A systematic review of research literature on unclassified tailings paste backfill technology at home and abroad was conducted, and the research progress was summarized from multiple dimensions such as technological development history, paste thickening and preparation, pipeline transportation, and engineering application cases. Research shows that the research on backfill technology at home and abroad has gone through four evolutionary stages: dry backfill, hydraulic backfill, cemented backfill, and paste backfill. Currently, paste backfill technology is developing rapidly in the direction of "high concentration, large flow rate, intelligentization, and green development". The core achievements of the complete technology system for paste backfill were systematically discussed, including the low-energy flocculation of tailings slurry synergistic with high-concentration stable sand discharge technology, the high-quality preparation of multi-component material paste synergistic with reliable transportation technology, the pressure-regulated paste transportation and backfill safety control technology in ultra-deep shafts, as well as the independent research and development progress of key backfill equipment such as paste storage thickeners, high-speed flexible mixers, and cone-valve industrial backfill pumps. Through engineering cases such as the Shizhuyuan polymetallic mine, the successful application effects of the technology and equipment in different mines were demonstrated. Combined with current technological research and future development trends, the development path of paste backfill technology towards deep underground and green development was prospected, aiming to provide technical reference and decision-making support for the green and low-carbon transition of the mining industry in China.

The treatment of goafs is crucial for residual ore recovery. Selecting a backfill with reasonable strength to treat the goaf, so as to prevent the premature instability of pillars caused by stress concentration, is a key prerequisite for ensuring the safe and economic recovery of subsequent pillars. To address the problems of early left-over goafs and residual pillars in a certain fluorite mine, the short-hole shrinkage subsequent filling mining method suitable for medium-thick ore bodies was adopted to recover the pillars. Furthermore, the strength range of the backfill was preliminarily determined to be 1.06-2.08 MPa through theoretical empirical formulas, and then the final backfill strength was determined in combination with numerical simulation. The results show that when the backfill strength is 2.0 MPa, it can effectively alleviate the stress concentration phenomenon at the top and bottom of the backfill, reduce the pressure relief range of the side walls on both sides, ensure that the displacement and deformation of the backfill on the side walls of the pillar stope do not exceed 50 mm, and simultaneously avoid the large-area plastic zone penetration between the backfill and the pillar. After the goafs were treated with backfill with a strength of 2.0 MPa at the site, the cumulative volume of filled goafs is approximately 337 339 m³. A total of 83 000 t of residual roof and sill pillar ores are recovered, and the total profit reaches 36.88 million yuan. While ensuring safe underground recovery, the economic benefits of the mine are significantly improved.

To address the time-varying problem of rheological parameters caused by cementitious material hydration during long-distance transportation of mine backfill slurry, rheological tests of backfill slurry were conducted at a constant temperature of 25 ℃, and the abrupt change law of the yield stress of cement-based backfill slurry at key nodes was revealed. A modified Darcy-Weisbach equation was derived, and an innovative design method for backfill pipelines based on time-varying rheology and sectional resistance was proposed, which was then validated through a 5 km backfill pipeline engineering case. The results are concluded as follows. Firstly, the higher the cement-sand ratio is, the earlier the abrupt change node appears, and the initial yield stress of backfill slurry with a high cement-sand ratio is higher than that with a low cement-sand ratio. Secondly, as the hydration time extends, excessive hydration products cause particle agglomeration and densification of the microstructure, leading to an increase in the plastic viscosity of the slurry and a significant increase in pipeline friction resistance. The greater the cementitious material content, the more obvious the change. Thirdly, compared with the conservative scheme, the time-varying sectional design scheme improves the safety margin of the backfill pipeline by 28.5%, and reduces the pipeline cost and annual energy consumption cost by 22.6% and 26.1%, respectively. The research results can provide a theoretical basis for the construction of mine backfill systems.

To address the influence of blasting fragmentation size on the comprehensive production costs of drilling, blasting and loading haulage in underground metal mines, 15 large-diameter long-hole stopes of a lead-zinc mine were taken as the research object. Based on measured blasting fragmentation size and production costs over the same period, a coupling model of direct operation costs of drilling-blasting-loading haulage was constructed, and sensitivity analysis and field industrial tests were conducted for verification. The results show that the comprehensive cost first decreases and then increases with the increase in average fragmentation size. The economic optimal fragmentation size is 24.99 cm, corresponding to a comprehensive direct operation cost of 7.501 yuan/t. When the explosive consumption per unit and blast hole pattern parameters are respectively perturbed by ±10%, the optimal fragmentation size remains within the range of 23 to 27 cm, and it is recommended to take 25 cm as the design target for on-site blasting fragmentation. Compared with the reference fragmentation size of 32.2 cm predicted by the Kuznetsov model, the predicted cost corresponding to the model's optimal fragmentation size is reduced by 3.09%. The average fragmentation size in the field test is 246.5 mm, which is 10.1% lower than that of the comparison stope, and the proportion of blocks larger than 600 mm decreases from 10.33% to 5.48%. The model is applicable to mines with similar geological conditions, mining methods, and equipment configurations, and its results only reflect the direct operation costs of the three processes of drilling, blasting and loading haulage.

Due to changes in physical and mechanical properties, the frozen rock mass in cold regions has a significant impact on rock fragmentation by blasting. To investigate the influence law of rock freezing temperature on the fragmentation characteristics of bench blasting, taking an open-pit mine in a cold region as the engineering background, a numerical simulation study on blasting was conducted using ANSYS/LS-DYNA software. Taking the normal temperature rock mass condition as a comparison, the influences of different freezing temperatures on blasting stress wave propagation, blasting crack distribution range, and rock mass damage and failure were mainly analyzed. The research results show that in terms of stress wave propagation, freezing temperature has a significant impact on the dynamic response of rock. As the rock freezing temperature decreases from 20℃ (normal temperature) to -30 ℃, the elastic modulus of the rock mass increases, and the stress wave velocity and stress amplitude increase by 27% and 20%, respectively, while the attenuation coefficient shows no significant change. In terms of blasting crack distribution range and rock mass damage and failure, the mechanical parameters of bench frozen rock mass in cold regions, such as compressive strength and elastic modulus, are higher, which has an adverse effect on rock fragmentation by blasting. In particular, for the frozen rock mass at -30 ℃ in the upper part of the bench, compared with the normal-temperature rock mass, its fracture zone range decreases by about 29%, and the reduction in damage and failure volume reaches 22%. The research results can provide a theoretical basis for the optimization of bench blasting parameters of frozen rock mass in cold region open-pit mines.

With the in-depth advancement of the "Dual-Carbon" strategic goals and the development of new-quality productive forces, intelligent new-energy equipment has been increasingly widely applied in the mining industry, and the lithium-ion battery power supply has become an important power source for mine equipment. However, due to the complex working environment and limited space in mines, the risk of thermal runaway of lithium battery and its secondary disasters is significantly increased. To effectively prevent and control the disaster risk of lithium-ion battery power supply in mines, the disaster chain of "inducement - thermal runaway warning - secondary disaster" was reviewed. The results show that the lithium-ion battery power supply has formed a relatively systematic thermal runaway inducement analysis and monitoring and early warning system in the field of commercial and energy storage, but the research on the thermal runaway evolution mechanism in the complex environment of mines is still not deep enough. The early warning technology of mine thermal runaway should be further developed in the direction of multi-source information fusion and intelligence, and the coordination construction of active early warning and passive protection measures should be strengthened. The research results can provide support for the safe application of lithium-ion battery power supply in underground mines.

With the continuous increase in mining depth and the growing complexity of operating environments, the traditional safety governance model for non-coal mines is facing severe challenges. It is urgent to promote the transformation of this safety governance model toward pre-incident prevention. The evolution context of safety production policies for non-coal mines in China, which have shifted from "post-incident accountability" to "source control" and further to "systematic governance", was systematically sorted out, and a four-dimensional analytical framework of "access control-process supervision-transformation and upgrading-supervision and law enforcement" was constructed. The research shows that the current policy system has effectively promoted the optimization of industrial structure and has improved the safety situation. The number of accidents and fatalities in 2025 decreased by 49.0% and 38.3% respectively compared with 2018. However, in the face of new scenarios such as deep mine mining and intelligent construction, the policy system still has structural contradictions including lagging regulations and standards, implementation obstacles at the grassroots level, and insufficient technical support. Drawing on the experience of safety risk management in Australia, the United States and Canada, this paper proposes that China should establish a five-dimensional optimization path comprising improved legal frameworks, intelligent upgrading, collaborative governance, market-oriented incentives and accountability implementation to address the "last-kilometer" problem in policy implementation. In the future, in-depth research should be conducted on policy adaptation in cutting-edge fields, assessment of policy synergistic effects, and research on region-differentiated strategies, so as to provide institutional guarantees for the development of new-quality productive forces in the mining industry.

Lead-zinc ore is one of the important metallic mineral deposits in Hunan Province and has exerted significant environmental impacts on surface water and groundwater in the surrounding mining area. To investigate the influence of mining activities on surrounding water bodies, 16 groups of surface water samples and 12 groups of groundwater samples were collected. Piper trilinear diagram, Gibbs diagram, ion ratio analysis, chloro-alkaline index, factor analysis, correlation analysis, mineral saturation index, and geostatistical spatial analysis (Kriging interpolation, distance decay fitting, and directional variogram) were employed to analyze the hydrochemical characteristics and genesis of water quality in the study area. The results show that the water bodies in the study area are generally weakly alkaline, with Total Dissolved Solids (TDS) mass concentration ranging from 70 to 1 190 mg/L. The groundwater is dominated by SO₄-Ca and HCO₃-Ca·Mg hydrochemical types. The hydrochemical evolution is dominated by weathering dissolution of carbonate rocks, supplemented by forward cation exchange dominated by silicate rocks. The coupling of sulfide mineral oxidation and carbonate rock neutralization is the formation mechanism of the high , high TDS, and near-neutral pH characteristics of mine water. The influence of lead-zinc ore dissolution on different groundwater flow systems differs significantly, exhibiting strong spatial differentiation of hydrochemical characteristics. The water mineralization degree score (MDS) is centered on the mining core area and decays exponentially with distance, decreasing to 40% of the initial amplitude within 1 km and approaching the regional background level within 2-3 km. The influence width perpendicular to the groundwater flow direction is approximately 1.6 km. The impact of mining activities on surrounding water bodies shows a radial pattern, with stronger effects in water bodies closer to the mining area. The research results can provide a basis for water inrush risk warning, water pollution prevention and control, and rational utilization of mine water resources.

To reveal the failure mechanism and dynamic catastrophe evolution law of pillar group and solve the problem of large-area goaf collapse caused by pillar group instability in gently inclined orebody mining, the PFC2D numerical simulation method was adopted to conduct parallel uniaxial loading test simulations on three specimens with different strengths. The failure force chain and crack evolution characteristics of multiple pillars were analyzed, the chain instability mechanism of multiple pillars was revealed, and a key pillar theory based on trigger pillars, chain-breaking pillars, and critical pillars as well as a key pillar identification method were proposed. The method was verified by using the Fetr6 chromite mine case. The research shows that the instability of pillar group follows the law of chain failure, and pillars fail sequentially in the order of their strength from low to high. The trigger pillar is the weakest pillar that fails first during the chain failure process. The chain-breaking pillar can stop the propagation of chain failure in the pillar group. The critical pillar determines the critical state of overall instability of the pillar group system. The proposed key pillar identification method can accurately identify key pillars in the chain dynamic instability process. Verified by the Fetr6 chromite mine case, when the pillar bearing area loss rate reaches 0.523, the pillar group system can be judged as overall unstable. The research results provide a theoretical basis and technical support for the stability assessment and disaster prevention and control of pillar group in mines.

To address the local instability problem of the talc-bearing soft rock slope on the east side of the M Pit in an open-pit copper-cobalt mine, engineering geological investigation, three-dimensional numerical simulation (FLAC3D), and limit equilibrium analysis (Slide 2D/3D) were comprehensively employed to investigate and analyze the stability of the talc-bearing slope. The slope instability mechanism was revealed and a corresponding treatment scheme was proposed. The research results indicate that the east slope of the M Pit is in a stable state as a whole, but the slope section with talc exposure at +1 335 m to +1 380 m is in an unstable state. The talc layer has low strength and is prone to slaking upon contact with water, constituting a dominant weak interlayer. The shear strength of its interface with the surrounding rock decreases significantly, forming a potential sliding surface and leading to a deep shear-shallow tension failure mode of the talc-bearing slope. Accordingly, a comprehensive treatment scheme of "7% cement-treated soil replacement + steel-plastic geogrid + blind ditch drainage at the slope toe" was proposed and implemented. Nearly one year of on-site monitoring shows that the slope displacement curve is generally stable and the treatment effect is remarkable. The established treatment technology framework of "instability mechanism analysis-quantitative stability evaluation-treatment scheme design-treatment effect monitoring feedback" can provide a theoretical basis and engineering reference for the treatment of slopes containing weak interlayers such as talc.

The safety and stability of the tailings pond are crucial for ensuring underground collaborative mining. To investigate the influence of the tailings pond on mine engineering, a three-dimensional geological-structural model of a certain lithium-containing porcelain stone mine was established using the coupled GTS/NX and FLAC3D modeling method, and the optimization of the sealing length of the +225 m adit within the inundation line of the tailings pond and the analysis of the influence of the tailings pond on the stability of ore body mining were conducted. The results show that significant changes are observed in the stress, displacement, and plastic zone of the surrounding rock of the adit under different sealing lengths. The optimal sealing length is determined to be 105 m, at which the tensile stress at the adit opening is reduced to 0.20 MPa, the displacement deformation is 8 mm, and no plastic zone is generated in the sealing area, so that the water pressure of the tailings pond can be effectively controlled and the stability of the adit can be ensured. It is found from the comparison of the mechanical responses of ore body mining under conditions with and without the tailings pond that, when the tailings pond is present, the tensile stress of the roof of the mined-out area is reduced from 0.87 MPa (under the condition without the pond) to 0.84 MPa, the displacement is reduced from 15 mm to 14 mm, and the plastic zone of the mined-out area is not connected with the surface or the plastic zone of the tailings pond. Therefore, under the current engineering and geological conditions, the tailings pond is not the dominant factor for the stability of ore body mining. However, stress and deformation monitoring still needs to be strengthened during the subsequent mining process. The research results can provide method reference and data support for the collaborative safety design and assessment of tailings ponds and underground mining in similar mines.

Tailings ponds are mostly located in mountainous areas with scarce hydrological data, and the rational formula method is widely adopted for their design flood derivation, in which the determination of the confluence parameter m value directly affects the engineering scale, safety and reliability, and economic benefits. However, in actual engineering design and safety evaluation, the determination of the m value has often not received enough attention, leading to deviations in calculation results and affecting the rationality and safety of the project. Confluence parameters not only depend on natural conditions such as the underlying surface of the basin, but also have a dynamic coupling relationship and nonlinear feedback effect with elements such as peak discharge and confluence duration. According to the Regulation for calculating design flood of water resources and hydropower projects and the Hydrological Manual of Qinghai Province, the m value ranges were determined respectively, and the response laws of confluence time, peak discharge, and average flow velocity to the variation of the m value were systematically investigated. The results shows that the confluence parameter has a significant regulatory effect on peak flood discharge, and this response mechanism has not received sufficient attention. Given that the catchment areas of tailings ponds are mostly small-scale watersheds, it is recommended to scientifically determine the m value based on systematic analysis of topographic and geomorphological characteristics, vegetation coverage characteristics, and field investigations, following the concept of "localized adaptation and multidimensional verification", so as to improve the scientificity and reliability for the design and construction of tailings ponds.

Intelligent upgrading of mining equipment is the core pathway to address the safety and efficiency bottlenecks in deep mining and to achieve high-quality development of the mining industry. The development process of Changsha Institute of Mining Research Co., Ltd from technology introduction to independent innovation was systematically reviewed, and the complete intelligent equipment system covering shaft construction, rock drilling and excavation, and underground auxiliary operations was introduced. Key technological breakthroughs in CKT series raise boring machines, CKY/CKQ series full-spectrum mining equipment, and intelligent auxiliary equipment integrating AI vision and robotics were highlighted. In response to the common industry shortcomings, such as insufficient adaptability of equipment and difficulties in multi-machine coordination under deep and complex working conditions, a cutting-edge technology layout covering intelligent sensing, digital twin, and cluster coordination was proposed, and the strategic path for transformation from a single equipment supplier to a full-scenario intelligent mining service provider was clarified. The research and development practices of Changsha Institute of Mining Research Co., Ltd have provided an important model for China to break away the dependence on imported high-end mining equipment and ensure the independent controllability of mining equipment.

To meet the requirements for improving the load capacity and operational efficiency of the auxiliary shaft hoisting system in mines, the extra-large cage for modern mine auxiliary shaft was designed and studied based on the finite element method. The mechanical performance of the overall cage and its key components under a fully loaded condition of 150 t was systematically evaluated by establishing a three-dimensional solid model and conducting static simulation analysis using the ANSYS Workbench platform. The results indicate that the maximum equivalent stress of the overall cage is 47.749 MPa, and the overall safety factor, calculated based on the yield strength of Q690 steel, is 14.45. The safety factors of the main beams, columns, and tail beam are 16.3, 19.9, and 20.4, respectively, all of which are above the design control value n_min=13 and are in compliance with the safety regulations. The structural stress distribution is reasonable, and the deformation is controllable. It is demonstrated by engineering application that the overall transportation capacity and operational safety of cage for large-scale equipment can be improved by the design, and that a systematic design method from design constraints, load transfer paths, global structural response, local strength verification to engineering validation is established, and a reference is provided for the structural design of extra-large hoisting cages in deep mines.

Blasting fragmentation is regarded as an important basis for evaluating blasting effects and optimizing blasting parameters, and the subsequent operations such as shoveling, transportation, and crushing and grinding are directly affected. With blasting fragmentation prediction and post-blast fragmentation recognition as the main line, the formation mechanism and main influencing factors of blasting fragments were introduced. Prediction methods, including empirical and semi-empirical models, mechanism analysis and numerical simulation, statistical learning and artificial intelligence, as well as recognition techniques, including traditional detection, 2D image recognition, 3D point cloud, and photogrammetry, were systematically reviewed. The characteristics of different methods in terms of prediction accuracy, mechanism interpretability, computational efficiency, field applicability, and implementation cost were compared. The analysis results show that existing methods have their own advantages and applicable boundaries, and that accurate prediction, stable recognition, and low-cost application are still difficult to achieve simultaneously under complex mine conditions. It is concluded through analysis that current research is mainly constrained by the lack of model interpretability and cross-scene generalization ability, the vulnerability of recognition results to field environments, and the underdevelopment of data standards and engineering application systems. The synergy among mechanistic models, data-driven methods, and multi-source recognition technologies should be strengthened in future research, with emphasis placed on real-time fragmentation prediction based on multi-source data fusion, precise recognition and control in complex environments, and coordinated optimization of intelligent blasting equipment and prediction models, so that support can be provided for establishing a closed-loop control system comprising pre-blast prediction, post-blast recognition, effect evaluation, and parameter feedback.

To address the issue that existing platforms focus on equipment access, data aggregation, and business system integration, a five-layer architecture of "object perception - network data - model decision-making - control execution - business management" was constructed. Compared with the typical five-layer technology stack of "field layer - edge layer - IaaS - PaaS - SaaS" and the organizational architecture of ten major business systems, this framework further clarified three types of operating time scales, seven closed-loop links, and two types of feedback channels for production and quality. Analysis shows that intelligent mines are evolving from single-machine automation to cross-process and multi-equipment collaboration, but overall they are still dominated by local closed loops, and the feedback of product quality information to upstream production is insufficient; inconsistent data semantics, insufficient model generalization ability, non-standardized task interfaces, and imperfect cross-equipment safety interlocking remain the main problems restricting large-scale application. To this end, based on a unified object model and digital thread, it was proposed to advance mechanism-data fusion, cloud-edge-terminal collaboration, and a task-driven safe autonomous closed loop, in order to promote the whole industry chain collaboration of planning, production, control, and product quality.

The engineering basis for introducing large models in mines is established through the long-term accumulation of digitalization, automation, and specialized models. However, the integrated understanding of complex risks by safety systems is still hindered by fragmented data, isolated knowledge, and the disconnection between early warning and response. Based on the technological evolution of intelligent mines and engineering practices, the prior foundations, including platform interfaces, specialized tools, multi-source safety evidence, and domain knowledge, were examined, and the differentiated requirements of high-risk scenarios such as goaf, tailings pond, and hoisting system were analyzed. The study suggests that mine safety large models should not replace control systems and on-site responsibility chains, but rather serve as a "cognition-and-proposal layer" to participate in multimodal understanding, risk assessment, mechanism interpretation, and solution generation. Its outputs should be checked by data evidence, professional rules, mechanistic models, protection interlocks, and human review, and routes for rejection and safe fallback shall be retained. Data governance and knowledge base, cognition and proposal, constraint checking, on-site response, and outcome feedback are incorporated into the same accountability chain by the resulting layered collaborative architecture. The engineering value of mine safety large models should possess interpretability, traceability, verifiability, and fallback capability.

To address the problem that the planetary gears in the gearbox of mine hoists operate under long-term low-speed, heavy-load, and alternating-load conditions, and that single acoustic or vibration signals are easily affected by propagation path, measuring point position and background noise, leading to insufficient feature extraction of fault signals, a fault diagnosis method based on point-symmetric acoustic-vibration fusion and ARFM-Net is proposed. Firstly, an acoustic-vibration synchronous acquisition test system for the planetary gearbox was established, and the differences and complementarities of acoustic and vibration signals in expressing fault information were analyzed. Then, Variational Mode Decomposition (VMD) and Feature Mode Decomposition (FMD) were adopted to process the two types of signals according to their respective characteristics. The effective modal components containing major fault information were selected and mapped into a polar coordinate feature map by using the point-symmetric graph method to realize the image-level fusion of multi-scale acoustic-vibration information. Finally, an ARFM-Net diagnostic network composed of Adaptive RepLK and Frequency-aware Moga was constructed. An adaptive large-kernel convolution was used to enhance the extraction capability of local fault details and long-range texture correlations, and a frequency-aware multi-level gated feature aggregation was used to highlight fault-related information and suppress background redundant textures. The experimental results of five types of planetary gear fault states show that the average recognition accuracy of the proposed method reaches 98.72%, and it shows favorable diagnostic performance in both comparative and ablation tests. This method can effectively improve the fault-recognition accuracy of planetary gears in the gearbox of hoists under complex working conditions, and provide technical support for intelligent monitoring of operation status of hoist transmission systems in mines.

To improve the adaptability of the energy management strategy for mining hybrid loaders to variable cyclic operating conditions, an energy management strategy integrating Deep Q-Network (DQN) and Transfer Learning (TL), named DQN-TL, was proposed. The core idea of this method is that the inherent cyclic characteristic of the loader's V-type operation is utilized, and such repeatable patterns are regarded as extractable and reusable high-value knowledge. First, DQN is utilized by the strategy to learn and distill the general control laws for typical operating cycles in a data-rich source domain. Then, through parameter transfer, the relevant law features are introduced into the target domain model, which is then fine-tuned with a small amount of data to adapt to specific operating conditions. Finally, by establishing the intrinsic relationship between the structured cyclic characteristics of the operating conditions and transfer learning, the adaptability to complex underground conditions is improved. The simulation results show that, compared with a rule-based control strategy, the fuel consumption rate of the proposed DQN-TL strategy is 189.21 g/(kW·h), the fuel economy is improved by 13.8%, its power allocation strategy is closer to the global optimal solution, and good potential for engineering application is demonstrated.

Aiming at the problem of difficult multi-protocol wireless access caused by the heterogeneous protocols of underground intelligent mine equipment and the harsh communication environments, a VxLAN over 5G multi-protocol transparent access scheme suitable for mine scenarios was proposed based on the special channel constraints and 5G private network transmission technology in underground mines. A unified two-layer virtual tunnel was constructed over the 5G network to directly encapsulate and transmit raw Ethernet frames of mainstream industrial protocols such as EtherCAT and PROFINET, without secondary modification to the protocol stacks of existing mine equipment, which greatly reduces the cost of upgrading and reconstructing underground communication networks. The actual verification was carried out by building a simulated mine underground 5G private network test platform. The results show that the end-to-end average latency of EtherCAT protocol data is 22.508 ms, and the maximum latency is 44.794 ms. The average latency of PROFINET protocol data is 21.007 ms and the maximum latency is 44.477 ms. The latency of both protocol data can stably meet the real-time requirements for scenarios such as condition monitoring of fixed underground mine equipment (latency≤100 ms) and conventional remote control of auxiliary operation robots (latency≤50 ms). However, for strong real-time control scenarios such as mining faces that require millisecond-level synchronization accuracy (latency≤1 ms), the current scheme cannot support it, and needs to be further optimized in combination with 5G URLLC enhancement features. The research results can provide a feasible technical path with low transformation cost and high versatility for the wireless access of multi-protocol heterogeneous equipment in non-coal mines, and provide engineering reference for the refined service bearer of 5G private network in intelligent mines.

To address prominent problems prevalent at the regulatory end of the dual-prevention mechanism, including data island, risk identification relying on fixed thresholds, and difficulties in penetrating enterprise underlying data, and to explore an effective path for digital-intelligent transformation of the regulatory end, a four-level regulatory framework of "data penetration-early warning grading-closed-loop response-law-enforcement implementation" was proposed. By realizing real-time interconnection between enterprise underlying sensing data and the regulatory end, this framework conducts cross-verification on the authenticity of hazard investigation, rectification management and monitoring data using methods such as spatio-temporal behavior verification, spatial static verification, and data consistency verification. On this basis, an early warning comprehensive index model was established to realize dynamic risk identification and hierarchical response. At the same time, a differentiated and accurate law enforcement path based on a digital evidence chain and "dynamic safety risk profile" was explored to improve the timeliness and accuracy of supervision. The empirical results show that this model improves the efficiency of mine supervision by more than 60% and effectively reduces the frequency of accidents. It provides a scientific scheme for promoting the transformation of the dual prevention mechanism from formal operation to substantive operation, and provides a feasible path for building a more efficient and intelligent safety governance system for non-coal mines.

A comparison and selection study on comminution process schemes was conducted for a super-large low-grade porphyry copper-molybdenum mine in Inner Mongolia (with reserves of 950 million t, a design scale of 70 000 t/d, a Cu grade of 0.165%, and a Mo grade of 0.032%). The ore is moderately hard to hard and contains easily sliming minerals such as chlorite. Based on a systematic review of the application status of three types of comminution processes, including conventional comminution, semi-autogenous grinding (SAG), and high-pressure grinding rolls (HPGR), four schemes were proposed: the semi-autogenous grinding+ball mill+pebble crushing (SABC) process, 50 mm feed high-pressure grinding roll dry process, 30 mm feed high-pressure grinding roll dry process, and high-pressure grinding roll wet process. A technical and economic comparison was conducted from the aspects of equipment configuration, construction investment, energy consumption, operating costs, and total lifecycle costs. The results show that the SABC process has the lowest construction investment, which is 17.89% lower than that of the high-pressure grinding roll wet process, but has the highest production cost per ton of ore (23.55 yuan/t); the high-pressure grinding roll wet process has the lowest production cost per ton of ore (20.07 yuan/t), which is 17.34% lower than that of the SABC process. The high-pressure grinding roll wet process has the lowest present value of the total lifecycle cost, indicating its optimal long-term economic performance. Comprehensively considering the ore characteristics, equipment maturity, and total lifecycle economic performance, the high-pressure grinding roll comminution wet process of "cone medium crushing, closed-circuit screening, high-pressure grinding roll ultra-fine crushing, wet closed-circuit screening, and one-stage closed-circuit grinding" is recommended, to provide a reference for the selection of comminution processes for similar super-large low-grade copper-molybdenum mines.

To address the mining and separation difficulties of a low-grade zircon-hafnium placer in a coastal shallow sea of Hainan, a systematic process mineralogy study was conducted using analytical techniques such as X-ray diffraction, mineral liberation analyzer (MLA), and scanning electron microscopy-energy dispersive spectroscopy (SEM-EDS). The chemical composition of the ore, mineral composition, dissemination size of valuable minerals, and mineral occurrence characteristics were clarified. Furthermore, a feasibility study on pre-screening and centrifugal tailing discarding was carried out using a self-made LX-60 vertical cyclone centrifuge. The results indicate that the ZrO₂ grade of the raw ore is only 0.105 0%, the valuable mineral zircon accounts for only 0.14%, and the content of light gangues such as quartz and feldspar is up to 86.60%. 75% of zircons are monomer with particle size of 0.04-0.20 mm, showing a significant specific density difference from the gangues, which is suitable for gravity separation waste discarding. 25% of zircons contain quartz, feldspar, and ilmenite inclusions with particle sizes less than 0.05 mm. This part of fine inclusions is easy to cause metal loss. The dissemination size of ilmenite is coarser than that of zircon, and it is synchronously enriched during gravity separation. After pre-screening out the +0.15 mm coarse material, the undersize material undergoes one-stage roughing under the conditions of a centrifugal force of 80g, a backwash water flow rate of 4 L/min, and a feeding amount of 80 g, yielding a pre-selected rough concentrate with a ZrO₂ grade of 0.483 6% and an operation recovery of 95.50%. The total zirconium recovery of the combined screening-centrifugation process reaches 93.37%, and the total loss rate to the raw ore is only 6.63%. This process achieves the efficient discarding of massive light gangues and the pre-concentration of zircon minerals, demonstrates potential applicability for offshore in-situ pre-selection tailing discarding, and can provide technical support for the pre-selection tailing discarding of similar extremely low-grade zircon placers.

The recovery rate of micro-fine cassiterite in a complex polymetallic tin ore in Yunnan has long been low, seriously restricting the improvement of comprehensive beneficiation indicators. Through systematic analysis of tin loss characteristics, the causes of tin loss were identified. Analysis of tin loss characteristics shows that the disseminated grain size of cassiterite is mainly concentrated in the range of 5-40 μm, with the proportion of the -20 μm fraction exceeding 55%, and the dissociation degree of cassiterite is only 31.63% at the current grinding fineness. The selectivity of the three-stage series hydrocyclone desliming is poor, the metal distribution rate of +10 μm cassiterite in the desliming overflow reaches 63.62%, and the cumulative tin metal loss from the multi-stage hydrocyclone desliming operation is about 7%. Conventional mechanical agitation flotation machines have insufficient recovery efficiency for -10 μm ultra-fine cassiterite. Technical renovation tests were carried out to break through the technical bottleneck of efficient recovery of micro-fine cassiterite. The results show that by replacing the vertical stirred mill with an IsaMill for fine grinding, the content of the -38 μm fraction increases from 50% to 75%, and the dissociation degree of cassiterite increases from 46.37% to 84.21%. By replacing the three-stage hydrocyclone with a decanter centrifuge for one-stage desliming, the tin recovery rate in the underflow increases by 8.39 percentage points. By applying a microbubble flotation column to replace the conventional mechanical agitation flotation machine, along with simplifying the flowsheet structure, the tin recovery rate is increased by 7.98 percentage points compared with on-site production. Simultaneously, the flotation reagent regime is optimized by adopting the combined collectors YK-Sn, Sn-705, and Sn-F, which increases the recovery rate of the tin rougher concentrate by 6.03 percentage points compared with the original collectors. Through the systematic integration of fine grinding, high-efficiency desliming, enhanced flotation and reagent optimization, the problem of micro-fine cassiterite recovery can be effectively solved, and the pilot combined flowsheet has achieved good recovery results. The research results can provide a reference technical route for the efficient recovery of micro-fine cassiterite from similar complex polymetallic tin ores.

To address the problems of high titanaugite silicate content in ilmenite and the low recovery of the existing gravity separation process from a certain area in Xinjiang, a systematic experimental study on flotation was conducted. Under the conditions of a raw ore TiO₂ grade of 15.64% and an ilmenite distribution rate of 77.37%, using a composite binary organic acid as the titanaugite depressant and adopting the beneficiation process flowsheet of "desulfurization reverse flotation-ilmenite direct flotation", an ilmenite concentrate with a yield of 26.02%, a TiO₂ grade of 45.62%, and a recovery of 76.27% was obtained through the closed-circuit test. Compared with the existing gravity separation process, the TiO₂ recovery rate in ilmenite concentrate by high intensity magnetic separation-flotation process is increased by 8.1 precentage points, and the ilmenite content in the concentrate reached 91.22%, which achieved the effective flotation separation of ilmenite and titanaugite. Zeta potential measurements and adsorption tests indicate that the composite binary organic acid has a greater impact on the surface potential of titanaugite and a higher adsorption amount, which increases the floatability difference between titanaugite and ilmenite, and provides a basis for their flotation separation.

To solve the problems of difficult lithium-sodium separation and low product purity in conventional processes, a pilot study on lithium extraction from spent lithium battery black mass was carried out using the sulfation roasting-alkali leaching process, with the innovative introduction of bipolar membrane electrodialysis technology. The pilot study showed that under optimized process conditions (a volume ratio of black mass to concentrated sulfuric acid of 1∶0.65 and roasting at 540 ℃ for 3.5 h), the lithium leaching rate was stably above 99.38%, and 99.65% of the fluorine in the black mass entered the flue gas, resulting in an extremely low fluorine content in the leachate. The nickel-cobalt-manganese residue obtained after alkali leaching was subjected to two-stage countercurrent acid leaching, and the leaching rates of nickel, cobalt, and manganese all exceeded 99.2%, enabling efficient recovery. To address the high sodium content in the alkali leachate, bipolar membrane electrodialysis was employed to convert the purified lithium sulfate solution into a lithium hydroxide solution with a hydroxide ion concentration of 1.38 mol/L and dilute sulfuric acid with a hydrogen ion concentration of 1.48 mol/L. The lithium hydroxide solution was further purified by nanofiltration and evaporation crystallization to prepare a lithium hydroxide product that met battery-grade quality requirements. The process is simple and efficient, with an overall lithium recovery of approximately 97.9%. Preliminary cost estimation shows that the total processing cost per ton of lithium produced by this process is approximately 94 100 yuan, demonstrating both technical feasibility and economic competitiveness and providing a new approach for the comprehensive resource utilization of spent lithium batteries.

Against the background of intensified price fluctuations and an overall oscillating upward trend in the metal market, traditional decision-making based on a fixed cut-off grade is difficult to adapt to market changes, which is detrimental to the long-term development of mining enterprises. Based on this, the internal logic and economic mechanism between metal price fluctuation and dynamic adjustment of mining cut-off grade were studied. Using the imperialist competitive algorithm (ICA), combined with models such as break-even analysis, multi-metal grade conversion, and stochastic price prediction, a logical framework for the dynamic adjustment of cut-off grade integrating data perception, model decision-making, intelligent execution, and feedback optimization was constructed. Taking a copper polymetallic mine as an example, four scenarios, namely static average price, progressive price increase, stochastic price path, and two-way price-cost linkage, were set up for empirical analysis. The results show that the cut-off grade is significantly negatively correlated with the price level, and dynamic adjustment can effectively prolong the mine production life and increase the net present value; the optimal strategy under stochastic prices is more conservative than that in deterministic scenarios, which can effectively mitigate the downside risk of metal prices; the cost elasticity coefficient has a significant regulating effect on the cut-off grade, and an increase in the cost elasticity coefficient reduces profit volatility while simultaneously weakening profit elasticity during periods of price increases. In the future, mining enterprises should focus on the deep integration of geological, market price and production data, and utilize intelligent technologies to achieve a fundamental transformation from "static design" to "dynamic response", thereby enhancing overall risk resistance capacity and resource utilization efficiency.

To effectively manage and control mine costs and guide enterprises in reducing costs and increasing efficiency, a cost target system was established based on the complete cost. Using Analytic Hierarchy Process (AHP), statistical analysis, graphical methods and case studies, systematic comparison and analysis were conducted on the complete costs of four medium-sized lead-zinc mines both domestically and internationally. The following conclusions are drawn. Mining methods, development systems, and production scales are crucial for controlling complete cost of mines. The compositions of complete cost in descending order of proportion are production cost, taxes and surcharges, administrative expenses, operating expenses, and financial expenses. The complete cost of mines has a scale effect, the larger the production scale, the lower the unit cost. For projects of the same production scale, the complete cost of foreign mines is higher than that of domestic mines. For medium-sized lead-zinc mines, the complete cost of lead is in the range of 10 000 to 17 000 yuan/t, and the complete cost of zinc is in the range of 6 000 to 12 000 yuan/t. The research findings can provide a reference for cost control in the practical production of lead-zinc mines.