An early-warning method for rock failure based on Hurst exponent in acoustic emission/microseismic activity monitoring

Catastrophic event in geotechnical engineering such as rock failure could cause serious casualties and property losses. Aiming at proposing an early-warning method for rock failure, precursors based on Hurst exponent and acoustic emission/microseismic activity monitoring were investigated. Firstly, Hurst exponent was introduced to the geotechnical field to reflect the long memory and fractal texture in time series. A modified R/S method was proposed with overlapping subseries, so fewer data were required for the calculation of Hurst exponent. Secondly, Hurst exponent was proposed varying with time, and the generation of Hurst exponent series was constituted. Finally, the failure of rock was studied with Hurst exponent, assisted with acoustic emission monitoring in laboratory testing and microseismic activity monitoring in field testing. Results show that the start of Hurst exponent parameter decrease and acoustic emission parameter increase could be viewed as the early-warning point of instability. Moreover, the drop in Hurst exponent combined with significant microseismic events or magnitude or sudden increase of cumulative apparent volume is defined as the early-warning point of instability. Hurst exponent with acoustic emission/microseismic activity monitoring for early-warning of instability is applicable and effective. The proposed method could be applied in disaster preventions with an artificial intellect.