A Rockburst Proneness Evaluation Method Based on Multidimensional Cloud Model Improved by Control Variable Method and Rockburst Database

Abstract Rockbursts are common geological disasters in underground engineering, and rockburst proneness evaluation is an important research subject. In this study, a multidimensional cloud model was used to evaluate the level, control variable method establish 15 (MC) models. The key factors affecting accuracy of numerical characteristics, weight, normalization methods. optimal characteristics calculation determined, improved CRITIC (IC) weight optimised by introducing relative standard deviation quantisation coefficient. Six indexes were as input for model, including elastic deformation energy index Wet, maximum tangential stress σθMPa cavern, uniaxial compressive strength σcMPa, tensile στMPa, brittleness coefficient B1=σc/στ, σθ/σc. learn 271 groups complete cases, MC-IC established. performance proposed verified 8-fold cross-validation confusion matrix (precision, recall, F1). tested 20 cases from Jiangbian Hydropower Station, reaches 95%. results compared with three empirical criteria: four cloud-based methods, unsupervised learning supervised methods; established paper 93.33%. showed that had excellent evaluating can provide practical basis identifying hazard areas deep engineering.