Real time seismic signal processing using the ARMA model coefficients and an intelligent monitoring system

A new method has been presented for real time seismic signal processing. A fuzzy model is extracted for the background noise dynamics, using the ARMA model coefficients. A fuzzy rule base has been generated, and a fuzzy inference engine has been used to detect the variations in the nature of the background noise. The conventional envelope detection algorithm for on-set estimation, has been affected by the fuzzy inference engine to make it more flexible and robust, in the presence of a large amount of background noise. The fuzzy inference engine also serves as a proper indicator for sudden variations in the nature of the background noise. If the detected variation is due to the first arrival phase of a seismic event, then a higher order ARMA model is derived and its coefficients are used as the inputs to a trained neural network, for seismic classification. The experimental results are promising and there are some remarkable advantages over the previous methods.