This study aimed to investigate the formation mechanism of the mud-inclusion-type underground debris flows of natural caving underground mines. The characteristics of fine moraine particles flowing through the coarse-grained ore bed were used to analyze the formation process of mud inclusions in the caving ore bed through a physical model test. Based on the movement behavior of the mud inclusions of moraine in the caving ore bed, a formation-mechanism generalized model of underground debris flows with mud inclusions was established. The model was used to examine the formation mechanism of mud-inclusion-type underground debris flows in natural caving. The results showed that the fine moraine particles had good cross-flow characteristics in the process of drawing coarse-grained ore. The accumulation of fine moraine in the ore bed was a prerequisite for the formation of mud inclusions, and the fluid inclusions were formed by a mixture of the particles with the infiltrated water. When mud inclusions in moraine are affected by many factors, such as ore-drawing vibrations, blasting vibrations, and groundwater, the inclusions undergo multiple migration–stop–migration cycles, resulting in separation or fusion. However, the inclusions are released along the optimal random pore path to the outlet, forming a certain scale of underground debris flows accidents. The accuracy and reliability of the formation mechanism were verified through geophysical explorations based on the equivalent inverse flux transient electromagnetic method. This study not only broadens the research on debris flow, but also provides theoretical guidance for the prevention and control of underground debris flows.

The shear factor of ore drawing is an important factor affecting the formation of underground debris flows. The aim of this study was to investigate the effect of the mining shear factor on underground debris flows in natural caving. The research background was the underground debris flow in the Plan copper mine, and we analyzed the characteristics of the slurry material structure of the underground debris flow, as well as the influence of the ore-drawing shear factor on the formation mechanism of the underground debris flow. The results showed that the slurry of the underground debris flow in the Plan mine is both a pseudoplastic and thixotropic fluid. Shearing force induced in drawing deforms the slurry and decreases its viscosity with the increase in shear rate and time. The shear force produced by the flow of ore particles first produces shear action on the paste in the shear boundary region of the ore drawing, reducing the paste viscosity while increasing its fluidity. Consequently, the “activation” makes the paste flowable, which flows along with the bulk ore flowing through the drawing mouth. The continuous ore-drawing process continuously shears the new moraine slurry in the ore-drawing channel and continuously “activates” the moraine slurry in the ore-drawing channel. Finally, destructive underground debris flow accident of a certain scale occurs. To our knowledge, this study thoroughly investigated the effect of the ore-drawing shear factor on the formation mechanism of underground debris flows, which not only broadens the research field of debris flow but also covers the deficiency of systematic research on underground debris flows, providing theoretical guidance for the prevention and control of underground debris flows.

Due to the impacts on the ecological environment, production safety and the economic benefits of large open-pit water deposits, there is no longer a single drainage method within the pit that can be used to meet the need for further deep mining. Therefore, curtain grouting technology is proposed for use in the prevention and control of mine water. The flow control equation of slurry in vertical fissures under the influence of self-weight stress was derived, and a three-dimensional visualized laminated jointed rock grouting model was developed independently to study the slurry transport mechanism, diffusion pattern, and sealing mechanism, which verified the correctness of the control equation. Field trials of industrial curtain grouting were also carried out at the mines, and the effectiveness of curtain water blockage was tested using visual analysis, inspection hole detection, and geophysical exploration methods. The results showed that after grouting the curtain, water-conducting cracks in the formation can be filled well and sealed, and the curtain has a good water-blocking effect. A complete set of diversified value-taking grouting systems is obtained, which can ensure the development of a green, safe and sustainable mine in the future.

To effectively mitigate the hazards posed by the blast waves of rock mass caving on closed walls during the mining process, a stability prediction method based on a random forest (RF) algorithm is proposed, which is designed to automatically identify key parameters. A machine learning model is developed using the algorithm, and its performance is evaluated through accuracy, precision, recall, and F1-score metrics. The probabilistic model of the objective function is constructed using the grid search hyperparameter optimization method, allowing for the selection of the most favorable hyperparameters for evaluation. The initial prediction accuracy of the RF algorithm model is 94.6%, indicating a strong predictive capability. Further adjustments to the base classifier, maximum depth, minimum number of leaves, and minimum number of samples enhance the model’s performance, resulting in an improved prediction accuracy of 95.9%. Finally, the optimized model is applied to predict the stability of the closed walls in the actual project, and the results are consistent with the on-site situation. This demonstrates that the random forest-based stability prediction model effectively forecasts the stability of closed walls in the actual project.

The accurate prediction of roof stress in mined-out areas is crucial for ensuring mine safety. However, existing study methods often overlook the increasingly available image data and fail to balance the model predictive capability with interpretability. To address these issues, this study innovatively integrates 3D laser scanning image features into the prediction of roof stress in mined-out areas. Image features are extracted using pre-trained deep-learning models and combined with traditional geological parameters to construct multiple machine-learning models for prediction. The experimental results demonstrate that models incorporating image features significantly outperform traditional models that rely solely on geological parameters in terms of prediction accuracy, interpretability, and complexity, revealing the critical role of image features in stress prediction. Furthermore, the use of SHapley Additive exPlanations (SHAP) to interpret the random forest model uncovers new domain knowledge, such as the relationship between spatial patterns and stress concentration. This study theoretically validates the effectiveness of image data and effectively balances the predictive capability and interpretability of the model, facilitating knowledge discovery in the field. On a practical level, it also provides guidance for mine safety management.