Automatic picking method of microseismic signal first arrival time based on empirical wavelet transform

Abstract

For problems in the process of microseismic signal acquisition, there is a large number of interference signals, which leads to that the first arrival time of signal is difficult to pick up.A picking method combines empirical wavelet transform (EWT) and signal first arrival algorithm is proposed to pick up the first arrival time of microseismic signal.Firstly,the component signals are obtained by the EWT method, which is used to decompose the microseismic signals with the characteristics of adaptive and conquering modal aliasing. Then the correlation coefficient and variance contribution rate of each component signals are calculated, and the component signals within the threshold range are selected for reconstruction to reduce the interference of noise signals. Finally, the modified energy ratio (MER) and fractal-dimension (FD) are used to pick up the first arrival time for the reconstructed denoising signal. According to the experimental verification, the accuracy of signal first arrival time picking combined with the MER and FD method based on EWT is 0.59% and 0.36% respectively. But in terms of efficiency, the former is obviously higher than the latter. The experimental and analytical results show that the model is highly accurate in picking up first arrival time of the effective signals and has a strong applicability to the signals of different data quantities.