Aiming at the problems of uneven illumination, complex background and large difference in structural plane morphology of borehole images, a dual-encoder interactive attention enhanced fusion network ( DISA-FuseNet ) was proposed. Based on the UNet encoder-decoder architecture, the network adopted the dual-encoder structure of Swin Transformer and ResNet50 to capture global context information and local detail features simultaneously. The attention was weighted by the Statistical Aggregation Channel Attention ( SACA ) module to suppress the background noise and irrelevant information of the rock mass. The underlying features were inputted into the Adaptive Hierarchical Enhancement module ( AHE ) after splicing to enhance the network's ability to capture multi-scale structural plane features. Experimental verification was conducted based on the self-constructed borehole structural plane dataset, and DISA-FuseNet was compared with five models, namely UNet, Deeplabv3+, Trans-UNet, DeepCrack and CrackFormer. The results show that DISA-FuseNet outperforms all the other models in terms of mean intersection over union, mean pixel accuracy, recall, precision and F1-score, these indexes are increased by 8.44, 9.51, 8.52, 10.60 and 9.62 percentage points respectively compared with those of the baseline model UNet. Furthermore, a method for structural plane parameter extraction and 3D visualization is proposed, which enables the accurate extraction of structural plane parameters and the construction of an interactive 3D visualization model. The actual engineering case verifies that the method has good effectiveness and engineering applicability.

Aiming at the challenges posed by the spatiotemporal randomness of wind resources in large-scale wind power bases in mining areas to the design of high-power wind turbine generators, based on the measured data of a wind farm in a mining area of Inner Mongolia, the adaptive design and optimization research on the 6.0 MW permanent magnet wind turbine generator was conducted. Based on multi-dimensional wind measurement data mining, the joint probability distribution of wind speed, temperature and duration was quantified, a dynamic load spectrum model reflecting the actual wind conditions was established, and with the objectives of improving electromagnetic characteristics and reducing material costs, the coordinated optimal design of coil size and permanent magnet size was carried out. The electromagnetic scheme was verified by finite element simulation, and the design target of 98% efficiency was achieved. The test results show that the errors between the no-load and load characteristics of the two developed prototypes and the simulation results are all within 3%, which verifies the high accuracy of the model. The temperature rise test reveals the temperature field distribution law of the generator under complex operating conditions, and the thermal management and control strategy is formulated accordingly. The study forms a complete technical chain consisting of "data-driven modeling-multi-objective optimization-simulation verification-prototype testing", which not only provides solutions for specific wind farms, but also offers references for the engineering application of high-power permanent magnet wind turbine generators in large-scale wind power bases.

Chengchao Mining Company adopts the principle flotation flowsheet of bulk flotation followed by separation to recover valuable resources such as copper and sulfur from iron tailings, but the copper recovery rate is relatively low. Therefore, aiming at the bulk flotation concentrate, the effects of factors such as regulators, depressants, collectors and pulp mass concentration on the separation flotation performance of copper and sulfur were studied in detail through tests. The results show that the optimal separation flotation conditions are as follows: pulp mass concentration of 30%, activated carbon dosage of 120 g/t, quicklime dosage of 6 000 g/t, and Z-200 dosage of 30 g/t. Under these flotation conditions, by adopting the process flowsheet of one roughing, two cleanings, one scavenging and sequential return of middle ore, a copper concentrate with a Cu grade of 16.00% and a Cu recovery rate of 54.87% can be obtained. A sulfur concentrate with a sulfur grade of 35.10% and a sulfur recovery rate of 33.57% can be obtained. The copper content of the flotation tailings is 0.013% and the sulfur content is 6.89%, among which the copper content in sulfides is 0.006 3%. 97.97% of the sulfur element exists mainly in sulfates, which indicates that the adopted reagent system and flotation process flowsheet can fully recover copper and sulfur from iron tailings.

Coal flotation tailings, as a low-value product of coal preparation plant, is rich in potential resources such as kaolinite. The coal slime from Malan Coal Preparation Plant was used as an experimental object. The mineral composition and particle size distribution was analyzed, the influences of flotation time and reagent consumption on the recovery of middling coal slime were researched, as well as the influence of structure and operation parameters of hydrocyclone on classification of kaolinite. On this basis, a process experiment of three-stage cyclone classification and flotation combined process was carried out, and the combined process of "three-stage classification and coarse-fine flotation" was proposed to recover kaolinite. The results indicate that the main minerals in the sample are kaolinite and quartz. Under the conditions of 1 500 g/t of collector, 150 g/t of foaming agent, and 5 min of flotation time, the ash content of tail coal can reach 80.11%. Under a certain feeding pressure, the smaller the diameter of the cyclone, the more significant the enrichment effect on kaolinite. The combined process of "three-stage classification and coarse-fine flotation" can obtain kaolinite products with a yield of 20.26% and a grade of 78.88%, and middling coal slime products with a yield of 40.85% and ash content of 33.87%. At the same time, about 38.89% of coarse sand tailings (ash content of 82.45%) can be directly discarded. Research can provide new technological approaches for improving the quality and resource utilization of flotation tailings in coal preparation plants.

In the future, China will implement manned lunar landing missions, establish a permanent base on the lunar surface, and realize the exploitation and utilization of lunar mineral resources. To address the issue of determining bearing capacity of lunar soil during lunar surface construction, a method was proposed to use small-scale laboratory loading tests to deduce the allowable bearing capacity and settlement of lunar soil simulant under large-scale foundations was proposed. Firstly, plate loading tests were performed on lunar soil simulant with a compactness of 78% using circular loading plates with diameters of 6 cm, 8 cm, and 10 cm, respectively. The measured load-settlement (p-s) curves of the lunar soil simulant were obtained, and through curve fitting, the initial tangent modulus and ultimate bearing capacity were extracted. The tangent modulus method was adopted to establish the relationship between the tangent modulus of the lunar soil simulant and its initial tangent modulus, applied additional stress, as well as ultimate bearing capacity. Furthermore, based on the settlement calculation theory, the p-s curves under foundations of arbitrary sizes were calculated. Validation against the results of small-scale load tests demonstrated a good agreement between the calculated p-s curves and the measured ones. Secondly, the proposed method was employed to compute the p-s curves under large-scale foundations, and the allowable bearing capacity of large-scale foundations was comprehensively characterized by integrating the ultimate bearing capacity criterion and settlement control method. The analysis results of bearing capacity of lunar soil simulant under large-scale circular and square foundations indicates that with the increase in foundation size, the allowable bearing capacity of the lunar soil simulant under both types of loading plates increases. Additionally, the allowable bearing capacity of the lunar soil simulant under square loading plates is higher than that under circular ones of the same size. It can be inferred from the calculated p-s curves that the settlement of the lunar soil simulant corresponding to the determined allowable bearing capacity meets the engineering requirements. According to the standard size of circular loading plates used in conventional loading tests, the allowable bearing capacity of the lunar soil simulant employed in this study has been deduced to be 122 kPa. The proposed method for determining the allowable bearing capacity of lunar soil simulant provides a scientific approach and theoretical basis for the determination of lunar soil bearing capacity during the construction of lunar bases and the exploitation of lunar mineral resources.

Sediment plumes generated during the mining of polymetallic nodules have a significant impact on the seabed ecosystem. To explore the sediment resuspension and dispersion phenomena under the shear disturbance of mining vehicles, a simulation study was conducted based on a finite element analysis of Eulerian model. The plume dispersion under different roller velocities (0.1-0.9 m/s) and indentation depths (33 cm and 18 cm) was analyzed, and the effectiveness of plume shields was evaluated. Results indicate that when the roller velocity exceeds 0.3 m/s, the plume size and height significantly increase, with the maximum plume height reaching 1.4 m. As reducing the indentation depth, the average plume diffusion velocity and plume size noticeably decreases. Additionally, with the addition of a plume shield, the plume size reduces by about 51%, plume height decreases by 25%, and diffusion tendency reduces to 46%, indicating that the plume suppression devices such as shields can effectively mitigate the plume. The study can provide a reference for controlling plumes in the mining of deep-sea polymetallic nodules.

The stability of deep roadway is one of the keys and difficulties in deep mining. Taking the deep mining of a hard rock metal mine as the engineering background, according to the ground pressure stress test results, FLAC3D was used to construct the numerical model, and five representative altitudes were selected to quantitatively analyze the auxiliary ramp stability under different support schemes from four aspects of stress, displacement, strain and plastic zone. The results show that when the circular arch is supported by C30 shotcrete with a thickness of 100 mm, the auxiliary ramp can remain stable at elevations of 2 325 m or above, but it is unstable at elevations of 2 175 m and 1 950 m, so it needs to be strengthened. When the floor is laid with C30 cast concrete with a thickness of 300 mm, the floor within the elevation range of 2 850 m to 1 950 m shows unstable conditions. Three schemes of strengthening support were studied respectively: only increasing the shotcrete strength of the arc arch support body, only increasing the shotcrete support thickness, and extending the arc arch support body into a cylindrical support structure. The results indicate that none of these schemes can maintain the stability of the deep section of the auxiliary ramp. A joint support scheme of "cylindrical structure+sprayed steel fiber concrete+lagged floor building" was proposed. Calculation results has verified the feasibility of this scheme, which is recommended for the auxiliary ramp below the elevation of 2 325 m in the mine. The supporting body is cylindge-shaped CF35 sprayed steel fiber concrete with a thickness of 100 mm.

In view of the problem of high-temperature thermal hazard in the underground of Sanxin Gold-Copper Mine, a pre-cooling system was constructed based on the stratum cold storage cooling technology, utilizing the abandoned roadways in the mine's constant temperature zone for cold storage in winter and cold release in summer, which can achieve the overall cooling of underground airflow. Through theoretical calculations combined with the local geothermal conditions, a cost-effective comprehensive cooling scheme was finally formed. The Ventsim software was used to simulate and optimize the ventilation system to verify the feasibility of the selected scheme. The results show that the mine's pre-cooling line is selected as Line 1, with a ventilation resistance of 801.71 Pa, which can be matched with the K40-6No.18 fan to ensure efficient cold energy transmission. Before the optimization of the ventilation system, the temperature of the air path in the bottom level of the model in summer is 29.5 ℃, and after optimization, it drops to 25.3 ℃, with a reduction of 4.2 ℃, achieving a good cooling effect. Field application shows that after the optimization of stratum cold storage cooling, the average wet-bulb temperature of each measuring point in the deep level drops to 24.62 ℃, and the average wind speed increases to 1.27 m/s, and the temperature of the air path after cooling meets the regulatory requirement of being lower than 27 ℃.

As a transition scheme between open pit mining and underground mining, hanging wall ore mining is of great significance to ensure the seamless connection of mine in time, space and yield. The hanging wall ore body in the western area of Luodang mining area was taken as the research object. In view of the complex geological conditions that the ore body is located under the high slope and adjacent to the open-pit transportation system, the two-step strip upward horizontal layered filling mining method was adopted. FLAC3D numerical simulation software was used to simulate and analyze the stability of stope, slope and open-pit transportation system in the process of mining. The results show that when the mining method is used to mine the hanging wall ore body, the plastic zone range and the maximum displacement of the stope roof are 3.4 m and 6.6 mm, respectively, and the plastic zone range and the maximum displacement of the slope and open-pit transportation system are 9.5 m and 5.8 mm, respectively. The maximum tensile stress of the stope, slope and open-pit transportation system is 1.2 MPa, and the overall stability is good. In the process of hanging wall ore mining, there is a significant negative correlation between the stope-slope distance with the stability of the stope. The research can provide theoretical reference for the mining of hanging wall ore bodies in similar mines.

To improve the efficiency and accuracy of the mining and stripping plan for large-scale open-pit mines, a dynamic sorting of mining increment bodies(ACO-DP) integrating the ant colony optimization (ACO) algorithm and dynamic programming (DP) method was proposed. Combined with the distributed search mechanism of ACO, the state transition path selection strategy was reconstructed, which significantly reduced the computational complexity. Through its practical application in an open-pit quarry in Guangdong, the significant effects of the ACO-DP method in improving the efficiency of mining and stripping plan, optimizing resource allocation and enhancing economic benefits were verified. The results show that the ACO-DP method shortens the planning time of the quarry from 3 000 minutes to 6.45 minutes, with an efficiency improvement of 99.8%. The stripping ratio is optimized from 0.39 to 0.124, with a decrease of 68.2%, and the net present value gradually increases from 244.99 million yuan to a peak of 253.93 million yuan. The ACO-DP method can be integrated into a mine digital twin system, supporting real-time dynamic adjustment and multi-objective collaborative optimization, and providing a new idea for intelligent mine planning under the "double carbon" goal.