The occurrence of rockburst can cause significant disasters in underground rock engineering. It is crucial to predict and prevent rockburst in deep tunnels and mines. In this paper, the deficiencies of ensemble learning algorithms in rockburst prediction were investigated. Aiming at these shortages, a novel machine learning model, deep forest, was proposed to predict rockburst risk. The deep forest combines the characteristics of deep learning and ensemble models, which can solve complex problems. To develop the deep forest model for rockburst prediction, 329 real rockburst cases were collected to build a comprehensive database for intelligent analysis. Bayesian optimization was proposed to tune the hyperparameters of the deep forest. As a result, the deep forest model achieved 100% training accuracy and 92.4% testing accuracy, and it has more outstanding capability to forecast rockburst disasters compared to other widely used models (i.e., random forest, boosting tree models, neural network, support vector machine, etc.). The results of sensitivity analysis revealed the impact of variables on rockburst levels and the applicability of deep forest with a few input parameters. Eventually, real cases of rockburst in two gold mines, China, were used for validation purposes while the needed data sets were prepared by field observations and laboratory tests. The promoting results of the developed model during the validation phase confirm that it can be used with a high level of accuracy by practicing engineers for predicting rockburst occurrences.

Rock lithology recognition plays a fundamental role in geological survey research, mineral resource exploration, mining engineering, etc. However, the objectivity of researchers, rock variable natures, and tedious experimental processes make it difficult to ensure the accurate and effective identification of rock lithology. Additionally, multitype hybrid rock lithology identification is challenging, and few studies on this issue are available. In this paper, a novel multitype hybrid rock lithology detection method was proposed based on convolutional neural network (CNN), and neural network model compression technology was adopted to guarantee the model inference efficiency. Four fundamental single class rock datasets: sandstone, shale, monzogranite, and tuff were collected. At the same time, multitype hybrid rock lithologies datasets were obtained based on data augmentation method. The proposed model was then trained on multitype hybrid rock lithologies datasets. Besides, for comparison purposes, the other three algorithms, were trained and evaluated. Experimental results revealed that our method exhibited the best performance in terms of precision, recall, and efficiency compared with the other three algorithms. Furthermore, the inference time of the proposed model is twice as fast as the other three methods. It only needs 11 milliseconds for single image detection, making it possible to be applied to the industry by transforming the algorithm to an embedded hardware device or Android platform.

The problem of backfilling pipeline invalidation has become a bottleneck restricting the application and development of backfilling technology. This study applied the whale optimization algorithm and random forest (WOA–RF) to predict the invalidation risk of backfilling pipelines based on 59 datasets from actual mines. Eight influencing factors of backfilling pipeline invalidation risk were chosen as the input parameters of the WOA–RF model, and the risk level was selected as the output parameters of the WOA–RF model. Furthermore, random forest, decision tree, artificial neural network, k-nearest neighbor, and support vector machine models were also established according to the collected datasets. The prediction performance of the six classification models was compared. The evaluated results showed that the established WOA–RF hybrid model has the best prediction performance and the highest accuracy (0.917) compared to other models, with the highest kappa value (0.8846) and MCC value (0.8932). In addition, the performed sensitivity analysis showed that the deviation rate is the most important influencing factor, followed by the internal diameter of the pipeline. Eventually, the WOA–RF hybrid model was used to predict the failure risk level of the backfilling pipelines of nine actual mines in Sichuan, China. The field datasets were collected through field investigation, and engineering verification was carried out. The research results show that the WOA–RF hybrid model is reasonable and effective for backfilling pipeline invalidation risk, and it can provide a novel solution for backfilling pipeline invalidation, with good engineering practicability.

Accurate rock mass quality classification is crucial for the design and construction of underground projects. Traditional methods often rely on expert experience, introducing subjectivity, and struggle with complex geological conditions. Machine learning algorithms have improved this issue, but obtaining complete rock mass quality datasets is often difficult due to high cost and complex procedures. This study proposed a hybrid XGBoost model for predicting rock mass quality using incomplete datasets. The zebra optimization algorithm (ZOA) and Bayesian optimization (BO) were used to optimize the hyperparameters of the model. Data from various regions and types of underground engineering projects were utilized. Adaptive synthetic (ADASYN) oversampling addressed class imbalance. The model was evaluated using metrics including accuracy, Kappa, precision, recall, and F1-score. The ZOA-XGBoost model achieved an accuracy of 0.923 on the test set, demonstrating the best overall performance. Feature importance analysis and individual conditional expectation (ICE) plots highlighted the roles of RQD and UCS in predicting rock mass quality. The model’s robustness with incomplete data was verified by comparing its performance with other machine learning models on a dataset with missing values. The ZOA-XGBoost model outperformed other models, proving its reliability and effectiveness. This study provides an efficient and objective method for rock mass quality classification, offering significant value for engineering applications.

Dynamic fatigue of rocks under repeated cyclic impact is a nonconservative property, as surrounding rocks in real environments subjects them to variable impact disturbances, and the degree of damage varies under different energy level loads. To evaluate the dynamic response and fatigue damage characteristics of rocks under multi-level cyclic impacts, uniaxial cyclic impact tests were carried out on granite with various stress paths and energy levels using a modified split Hopkinson pressure bar. Dynamic deformation characteristics of specimens under different loading modes were investigated by introducing the deformation modulus of the loading stage. Evolution of macroscopic cracks during the impact process was investigated based on high-speed camera images, and the microscopic structure of damaged specimens was examined using SEM. In addition, cumulative energy dissipation was used to assess the damage of rocks. Results show that the deformation modulus of the loading stage, dynamic peak stress and strain of specimens increase with the impact energy, and the deformation modulus of the loading stage decreases as the damage level increases. Propagation rate of tensile cracks in rock was correlated with participation time of the higher energy level, which observed the following sequence: linearly decreasing > same > linearly increasing energy level, and cyclic loading of nonlinear energy level produced more tensile cracks and rock spalling than the same energy level. Compared with cyclic impacts of the same energy level, multi-level impacts form more microcracks and fatigue striations. The cumulative rate of specimen damage under the same energy change rate is as follows: linear decreasing > same > linear increasing loading. This provides a new case study for evaluating the dynamic damage, crushing efficiency and load-bearing capacity of rocks in real engineering environments.