Decision Tree Model for Rockburst Prediction Based on Microseismic Monitoring

Abstract

Rockburst is an extremely complex dynamic instability phenomenon for rock underground excavation. It is difficult to predict and evaluate the rank level of rockburst in practice. Microseismic monitoring technology has been adopted to obtain microseismic events of microcrack in rock mass for rockburst. The possibility of rockburst can be reflected by microseismic monitoring data. In this study, a decision tree was used to extract the knowledge of rockburst from microseismic monitoring data. The predictive model of rockburst was built based on microseismic monitoring data using a decision tree algorithm. The predictive results were compared with the real rank of rockburst. The relationship between rockburst and microseismic feature data was investigated using the developed decision tree model. The results show that the decision tree can extract the rockburst feature from the microseismic monitoring data. The rockburst is predictable based on microseismic monitoring data. The decision tree provides a feasible and promising approach to predict and evaluate rockburst.