Abstract:It is crucial to explore the characteristics of rock mass failure and provide specific early warning parameter values to ensure mine safety in the study of early warning for rock mass damage in mines. This article takes a metal mine as an example and establishes a ground pressure monitoring system based on microseismic monitoring technology. The principal component analysis method is used to determine the optimal statistical time window and analyze the spatiotemporal evolution characteristics of microseisms before and after the roof caving event combined with a large-scale roof caving event in the mine. The analysis shows that the center position of the microseismic event density cloud map is consistent with the location of the failure, and there is a precursor information that the number of microseisms continues to increase before the rock mass failure, and then rapidly decreases after reaching its peak. The microseismic differentiation degree was expressed by the microseismic quantity variance to represent the damage degree of rock mass. The microseismic differentiation rate index was obtained by using the Finite difference method. The research shows that the microseismic differentiation rate of rock mass failure process was divided into three stages: stable differentiation stage, accelerated differentiation stage and rock mass failure stage. The Gaussian Amp function was used to fit it, and the upper limit of the confidence interval of 0.306 with a confidence level of 0.99 was ultimately determined as the early warning value for differentiation rate. This study provides a theoretical basis for the establishment of a mine disaster warning system using microseismic monitoring, and has certain reference significance for mine safety production management.