Abstract: In this paper, by using the large-scaled coal & gas outburst simulation experiment system, as well as the developed materials in cavity opening with the property of high sealing performance and powerful rheological behavior, the automatic outburst process was simulated under the comprehensive action of vetical and horizontal stresses. Meanwhile, the whole experiment processes were monitored by microseismic system with high sensitivity installed on the cavity surface. The gas pressure within the outburst assembly cavity in each stage was monitored, and the dimension and shape of the outburst cavity were outlined. The analyzed frequency ranges from 0 to 2 000 Hz. By the optimization of wavelet primary function, the maximum decomposition of wavelet packet was determined. Then, the frequency characteristics in the four stages was acquired by the decomposition and reconstitution of signals based on Wavelet Packet Transform. The results show that the signal frequencies in the inoculation stage are in the high frequency range, while the signals in the excitation stage start to broaden and the low frequency signals appear. In addition, the frequency in the outburst stage is distributed in the full-band, while the signals transmit from the low frequency band to high frequency in the residual stage. The research results can provide reference to the early-warning of outburst by microseismic monitoring.

Key words: coal &, gas outburst simulation, automatic triggering, microseismic response, Wavelet Packet Transform, frequency characteristics