Recognition of mine microseismic signals based on FSWT time-frequency analysis

Typical microseismic and blast vibration signals of mine rock mass are studied by using the frequency slice wavelet transform (FSWT). First, FSWT is used to decompose the signals in time and frequency domains, and time-frequency characteristics of two kinds of waveforms are analyzed. Next, owing to that the frequency bands can be chosen arbitrarily through inverse transform of FSWT, different energy distribution characteristics of the two kinds of signals are studied by building 6 continuous sub-bands so as to get the reconstructed signals, then the sub-bands are divided more narrowly for deeper researches. The results show that the energy of the two kinds of signals is mainly distributed below 100 Hz in this mine, and the difference is as follows: the energy of rock mass microseimic signals is mainly concentrated in the band between 0~50 Hz, but for the blast vibration signals, the energy is concentrated more obviously in the band between 50~100 Hz and has a higher energy proportion than that of the rock mass microseimic signals over 100 Hz.