In response to the problem of large computation and slow detection speed of the current pedestrian detection algorithm in underground mines, a lightweight underground pedestrian detection model named MBE-YOLO was proposed based on YOLOv8n. Firstly, a lighter neck structure was designed based on the weighted bidirectional feature pyramid network, with the goal of efficiently fusing multi-scale information. Secondly, the detection head of the original network was redesigned using shared convolution, and the phantom convolution was introduced to obtain rich and effective feature representations, thus solving the problem of the huge number of parameters in the original detection head. Finally, considering the impact of lightweight model on accuracy, a lightweight hybrid local channel attention mechanism was added, which adaptively adjusts the importance of different regions at the global and local levels to improve accuracy. Experimental results show that compared with the benchmark model YOLOv8n, the mAP@0.5:0.95 of MBE-YOLO model is improved by 0.2 percentage points, the computational cost is reduced by 44%, the FPS is increased by 4.3 frames/s, the number of parameters is reduced by 46%, and the model size is reduced by 43%. Compared with the five mainstream models, this model has significant advantages in detection accuracy and lightweight, making it more suitable for scenarios that require real-time detection.

Aiming at the problems of low efficiency and difficulty in applying traditional fire detection algorithms to embedded devices with limited resources, a lightweight fire detection algorithm based on improved YOLOv8n was proposed. Firstly, in order to greatly reduce the number of model parameters, the C2f module in the YOLOv8n model was changed to the C2fGhost module. Secondly, in order to prevent the loss of important features after the lightweight of the model, strengthen the feature extraction ability and reduce the interference of irrelevant information in the image, the EMA attention mechanism was added to the Backbone part of the YOLOv8n model. Finally, in order to improve the generalization ability of the model, the CIoU loss function in the YOLOv8n model was replaced by WIoU. The experimental results show that the parameters of the improved model decrease by 30.1%, mAP50−95 increases by 0.6 percentage points, and the accuracy rate decreases by only 0.2 percentage points. The improved model can detect the underground fire in real time and efficiently, and the detection algorithm has reference value for mine safety management.

In response to the call of Zhejiang Province to comprehensively promote the digital construction of mines, it is committed to realizing the digital transformation and upgrading of mines and promoting the high-quality development of enterprises. Aiming at the system construction requirements of basic digitization in small and medium-sized open-pit mines, modern information technologies such as real three-dimensional, digital twin, the Internet of Things and artificial intelligence were introduced. The system architecture, key technologies and core modules were deeply analyzed and studied. A mine basic digital management system with cross-temporal, full-cycle and multi-domain data fusion was constructed and applied in many small and medium-sized open-pit mines in Linhai City and Wenling City, Zhejiang Province. Its powerful data fusion ability and intelligent management level have promoted the transformation of mining enterprises from "fragmented" management to "integrated" management, realized the transformation of data resources to data assets, and have guided the transformation of enterprises from traditional thinking to modern system thinking. At the same time, it also provides a reference for the digital and intelligent upgrading of small and medium-sized mines in China.

Mineral resources are an important material foundation for human survival and development. To strengthen the research on regulatory technology for cross-border mining in open-pit mines, high-resolution remote sensing technology was proposed for dynamic monitoring of cross-border mining activities. Three periods of high-resolution remote sensing images were selected, the change information of cross-border mining patterns in open-pit mines with provincial and municipal certification in Xiangtan City was extracted, and remote sensing dynamic monitoring research was conducted on cross-border mining in open-pit mines. A remote sensing interpretation marker library was established for the development of open-pit mines in the research area, a relatively complete set of dynamic monitoring technology methods was explored for cross-border mining in open-pit mines. The accuracy of suspected cross-border mining and mining development of the map spot accounts for 100% and 90.3%, respectively. It is found that, from 2016 to 2020, the number, map spots, and total area of cross-border open-pit mines with provincial and municipal certification in Xiangtan City show an overall decreasing trend. The research results indicate that high-resolution remote sensing technology has a strong practicality and high accuracy in the dynamic monitoring of cross-border mining in open-pit mines, and can effectively monitor the dynamic changes of cross-border mining. The research results can provide a technical support for the dynamic monitoring of cross-border mining in open-pit mines in Xiangtan City.

In order to solve the problems of low efficiency, high labor cost, and inability to meet the large-scale drilling requirements of traditional RTK-line hole layout methods, the application of UAV-LiDAR technology in open-pit mines was explored. Taking the Southern Open-pit Coal Mine as the research background, comparative experiments were conducted with traditional RTK-line hole layout method in terms of layout error, operational efficiency, environmental adaptability, and economy. The results of field testing and data analysis show that the UAV-LiDAR hole layout technology can reduce the hole layout error from 17.4 cm to 0.1 cm, improve the operation coverage ability by 200%, and save a cumulative operation cost of 1.931 million yuan within 10 years. Meanwhile, it has an outstanding applicability in dealing with complex terrain, high temperature environments, and night work. Furthermore, this technology can provide precise coordinates for the collaborative operation of intelligent drilling machines, hole inspection robots and other equipment, promote the development of intelligent blasting, and provide a key technical support for the intelligent process of mining, with broad application prospects.

A foreign high-sulfur hematite-limonite ore contains 5.68% of sulfur and 45.95% of iron. Iron mainly occurs in hematite-limonite, and sulfur mainly exists in the form of sulfide. In order to rationally develop and utilize the ore, based on the study of the properties of the raw ore, the process flow of grinding-flotation sulfur separation-high intensity magnetic separation of iron was tested. Under the condition of grinding fineness of −0.074 mm accounting for 80%, the sulfur concentrate with sulfur grade of 41.31% and sulfur recovery of 69.74% was obtained, and the iron concentrate with iron grade of 60.84% and iron recovery of 76.92% was obtained. The sulfur content of iron concentrate meets the requirements of quality standards, and the comprehensive recovery and utilization of high-sulfur hematite-limonite ore resources are realized.

The efficient and sustainable utilization of resources can be achieved by flotation recovery of high-quality carbon components in high-ash fine-grained coal slime. Taking high-ash fine-grained coal slime as the research object, by compounding octylphenol polyoxyethylene ether (OP-10) and polyethylene oxide (PEO) for pretreatment and comparing with single reagent pretreatment flotation, it was found that the two reagents greatly improved the floatability of coal slime through synergistic effect. On this basis, the mesoscopic mechanism was analyzed by laser particle size analyzer, contact angle measuring instrument, infrared spectrometer (FTIR), Zeta potentiometer and atomic force microscope (AFM). The results are as follows. Firstly, due to the competitive adsorption relationship between the two agents, PEO agents preferentially adsorbed and exposed the alkyl chain through polar forces on the surface of the coal sample to enhance partial hydrophobicity, while compressing the hydration layer on the surface of the slime and reducing the electrostatic repulsion between the interfaces, prompting the fine coal slime to begin to flocculate into small flocs. Secondly, after the completion of PEO reagent action, OP-10 reagent began to adsorb on the surface of flocs, compressing the electric double layer of coal slime and shielding the electronegative ions. Thirdly, the two ends of the structure of the agent were adsorbed by the polar force, so that the middle aromatic hydrocarbons were partially exposed to the surface of the coal sample, showing the most obvious hydrophobic modification effect. In general, the compound reagent adsorbs in the form of single-layer adsorption and double-layer adsorption through electrostatic force and polar force to enhance the hydrophobicity of coal samples, and the adhesion force is greatly reduced after the action of the compound reagent, and the flotation efficiency is greatly improved.

A novel biodegradable surfactant of sodium N-lauroyl sarcosinate (LS) was used for copper ion flotation. The effects of mass concentration ratio of LS to Cu(Ⅱ)(ρ_LS:ρ_Cu(Ⅱ)), pH value of solution and flotation time on the flotation effect of copper ion were explored by single factor condition tests, and the flotation conditions were optimized by Box-Behnken response surface method. Finally, the verification test was carried out based on the optimization results. The results of single factor test show that at the best condition with ρ_LS:ρ_Cu(Ⅱ) of 2.5, pH value of 8, and flotation time of 5 min, the removal rate of Cu(Ⅱ) reaches 97.97%. The verification test results show that the average removal rate of Cu(Ⅱ) in the three verification tests reaches 99.71% at the condition with ρ_LS:ρ_Cu(Ⅱ) of 2.5, pH of 9, and flotation time of 4 min, which is basically consistent with the prediction results obtained by software fitting. It indicates that the model has a good fitting degree and reliability. The morphology and elemental analysis show that the structure of LS-Cu complex is loose and stable. Compared with LS, the mass concentration of Cu(Ⅱ) in the complex increases from 0.00 to 5.15%, indicating that LS has a good adsorption capacity for Cu(Ⅱ).

A systematic study was conducted on the influence of preparation conditions of caustic starch on the reverse flotation desilication effect of a certain Anshan style hematite ore in Hebei province. Through single-factor variable tests and closed-circuit process test, the optimization combination scheme of key preparation parameters such as causticity ratio, causticity temperature, and causticity time was clarified. The results show that under the optimal conditions of a causticity ratio of 8 : 100, a causticity temperature of 90 ℃, and a causticity time of 30 min, the selective inhibition performance of caustic starch was significantly improved. In the closed-circuit test, the iron grades of iron concentrate and tailings were 67.02% and 20.65%, respectively, with an iron recovery rate of 87.63%. Infrared spectroscopy and Zeta potential analysis revealed the mechanism of changes in the molecular structure and adsorption properties of caustic starch. The results confirmed that after causticization treatment, starch can effectively expose its hydroxyl and carboxyl groups, which enhances the chemical adsorption ability of the inhibitor on the surface of hematite, thereby achieving selective inhibition of hematite. The research result provides a novel approach for the green and sustainable development of the reverse flotation desilication process for hematite.

In order to solve the technical bottleneck of low reaction activity and limited usage of IOT (iron ore tailings) in preparation of geopolymers, IOT-based geopolymers were prepared with high-silica IOT and low-silica IOT as the main raw materials, metakaolin as the correction material and water glass as the alkali activator. The thermal-chemical activation was used to activate the activity of IOT. The XRD and SEM/EDS were used to analyze the changes in mineral composition of IOT before and after activation, and the influence of thermal chemical activation at 500 ℃ and 700 ℃ on the microstructure and structure of geopolymers was studied. The results show that the reactivity of IOT after thermal-chemical activation is significantly increased, the geopolymers have dense internal structure and the compressive strength is increased significantly. The suitable thermal-chemical activation temperature is 500 ℃. At the same thermal-chemical activation temperature, IOT with a 80 μm sieve residue of 10% exhibit the highest activity and the geopolymers prepared by them have the highest compressive strength. Compared with low-silica IOT, high-silica IOT can obtain a higher reactivity after thermal-chemical activation, and the geopolymers prepared with high-silica IOT have a higher compressive strength. Thermal-chemical activation causes partial transformation of low activity minerals in IOT, forming highly active minerals such as soluble sodium silicate and metastable amorphous substances.