LAUSR

Abstract Rockburst hazards are generally recognized as one of the main types of geological hazards that affect the construction safety of underground engineering in areas with high ground stress. As such, rockburst prediction has become a common challenge for engineers worldwide. This paper presents a pioneering rockburst prediction method, referred to as the two-step comprehensive evaluation model, based on multiple empirical criteria. Five empirical criteria – the elastic energy index criterion, the Russenes criterion, the Tao criterion, the rock brittleness coefficient criterion, and the rock mass intact coefficient criterion were adopted to construct the primary evaluation model. The weight coefficients were subsequently assigned to the five empirical criteria according to the rough set theory. Finally, the two-step comprehensive evaluation model was constructed using the obtained weight coefficient and the technique for order preference by similarity to an ideal solution (TOPSIS). Twenty rockburst samples were collected to assess the validity of the two-step model. The results indicated that the lowest error rate decreased significantly from 45% (calculated with the primary evaluation model) to 25% (derived from the two-step model). The two-step model was compared with existing artificial intelligence methods and the results showed that the two-step model had better prediction performance in this study of rockburst cases. Thus, the presented model offers guidance for predicting underground rockburst hazards.