Research of Micro-seismic Signal Extraction Method Based on Rough Set

Abstract

Micro-seismic monitoring as a regional monitoring means to predict important dynamic disaster of the mine, meanwhile, micro-seismic signal with abundant components of spectrum and frequency bandwidth characteristics. Obtain sudden change time of micro-seismic abnormal signal and frequency components corresponding to sudden change time are the key issues in micro-seismic monitoring which to be solved urgently. In view of the variable precision rough set can not identify random rule which is supported by only a few examples, study a new hybrid model which combines rough set with Bayesian probability. In order to get more general and more reliable classification rule, make full use of intrinsic characteristic of micro-seismic monitoring knowledge system, more powerful in coping with noise data, derive a reliable and simple classification rule, it is more efficient to analyze a large number of micro-seismic data.