Revisiting rockburst predictive models for seismically active mines

Rockburst is a mining-induced seismic event characterized by a sudden explosion of rock due to the release of strain energy stored in rock mass, often occurring in high geo-stress and other unfavorable geological conditions. The prediction of the rockburst damage potential (RDP) is essential in mining as this phenomenon causes damage to excavations, leading to dire consequences that include economical losses, injuries, or casualties of miners. Despite the existence of a large number of rockburst predictive models and other empirical tools, a reliable prediction of RDP still remains challenging. Hence, the aim of this paper is to revisit some existing studies and propose practical RDP charts based on basic machine learning algorithms such as artificial neural network classifier. Historical records of rockburst compiled from Australian mines were employed for this purpose. Overall, the results of this study showed good consistency with the field data and outperformed those of some existing studies. It is concluded that the proposed charts could be used for the excavation vulnerability assessment and, therefore, assist in managing ground prone to rockburst in seismically active mines.