Multivariate acoustic emissions precursors of rockburst from the perspective of early warning

Rockburst precursors are critical for disaster warning, yet the complexity of rockburst has hindered the identification of a unified precursor. Furthermore, the influence of loading rates (LRs) on acoustic emission (AE) precursors in different rock types remains poorly understood. This study investigates the AE characteristics and early warning times of rockburst in slate and mica-schist under four LRs (0.05, 0.15, 0.25, and 0.5 MPa/s) using true triaxial unloading tests. The micro-crack state of the samples was evaluated using entropy, while critical slowing down (CSD) theory was applied to interpret AE precursors. The results reveal that as the LR increases, the rockburst stress of both rocks initially rises and then declines, with mica-schist exhibiting more severe damage and a higher dominance of tensile cracks. Notably, identifying rockburst precursors in mica-schist proved more challenging compared to slate. Among the methods tested, AE amplitude variance outperformed entropy in precursor identification. Additionally, the rockburst early warning time was found to be negatively correlated with the LR, with mica-schist consistently showing shorter warning times than slate. The CSD-derived precursor, due to its enhanced sensitivity, is recommended for early warning systems. These findings provide new insights into the role of LRs in rockburst dynamics and offer practical guidance for improving precursor identification and disaster mitigation strategies.