Classification forecasting research of rock burst intensity based on the BO-XGBoost-Cloud model

The rock burst belongs to a serious engineering geological disaster among the underground and deep resource exploitation engineering. The prediction of rock burst intensity has become an urgent problem to be addressed in underground engineering. In order to effectively predict the rock burst intensity grades, the four indicators were comprehensively selected in this paper to establish a rock burst evaluation indicator system. More specifically, the Bayesian optimization algorithm was employed to optimize the XGBoost machine learning algorithm to determine the weights of various evaluation indicators. Afterwards, the cloud model theory was introduced into the grade prediction study of rock burst intensity, as well as ultimately establishing a comprehensive evaluation model of rock burst intensity based on the BO-XGBoost-Cloud model. Consequently, the feasibility and validity of the model were examined with the assistance of on-site measured data and rock burst discrimination results among the relevant literature, which indicated that: the BO- XGBoost-Cloud model exhibited the highest accuracy in predicting rock burst intensity, significantly surpassing the extension evaluation method and the entropy-cloud model. Therefore, the BO-XGBoost-Could model was the optimal model for predicting rock burst intensity, which provided a creative forecasting method for the classification prediction of rock burst intensity in underground engineering.