Abstract: Rockburst prediction has become one of the critical issues that urgently need to be addressed in the fields of underground construction safety and geological disaster prevention. To fully consider the influence of various key factors on rockburst and improve the prediction accuracy of rockburst intensity, this paper describes a new collaborative prediction method based on primary and auxiliary models, grounded in the concept of multi-model stepwise predictions. Using this method, we have developed an intelligent software for rockburst intensity prediction and applied to the Nanning Pumped Storage Power Station to assess the rockburst tendency in its powerhouse and tunnel bifurcation sections. The results show our method achieves high prediction accuracy and agrees well with the actual conditions. The software is simple, practical, and easy for engineering technicians to use.

Key words: rockburst, prediction, machine learning, rockburst intensity