As the two important metric parameters of non-Gaussian and asymmetrical distribution time series,skewness and kurtosis
represent deviation degree from symmetrical distribution (symmetry) and distribution concentration (Gaussian) of the signal
distribution respectively.Compared with conventional seismic data,the energy of microseismic signals are weak,and SNR is
extremely low,it is difficult to get effective microseismic signals by using conventional signal processing methods directly.In view
of this,an improved time-varying skewness and kurtosis method was proposed according to the characteristics of microseismic
signals,which can be used to identify effective microseismic signals in strong interference background.Firstly,the time-varying
skewness or the kurtosis for local seismic data is obtained in different length sliding windows.Then,the difference of time-
varying skewness or kurtosis is achieved in the long-window and the short-window,and the maximum difference values are
corresponding to the location of the effective microseismic signals.The analysis of the theoretical model and actual data showed
that the improved time-varying skewness and kurtosis method can well eliminate the impact of asymmetry or non gaussian of
noise and highlight effective microseismic signals.
论文详情