Abstract: In order to obtain effective precursor information of instability failure of coal sample under unloading conditions, a fixed axial stress unloading confining stress experiment under different gas pressures was designed, and AE results of coal sample fracture process were obtained, and critical slowing down characteristics of AE time series under different gas pressures were calculated, so as to provide early warning of occurrence of coal and rock gas dynamic disasters. The results show that under unloading conditions, AE time series of gas-bearing coal has critical slowing characteristics. Sudden increase of variance and autocorrelation coefficient before coal sample failure can be used as precursor information to characterize failure of sample. Window length and lag step can affect critical slowing down characteristics. In contrast, variance is less affected by window length and lag step, and curve changes smoothly before precursor information appears.

Key words: unloading conditions, gas-bearing coal, acoustic emission (AE), critical slowing down characteristics, precursor information