Microseismic Signal Recognition Based On Phase Space Reconstruction

Micro-seismic monitoring technology plays an important role in evaluating the effect of fracturing.At the same time,micro-seismic data also has an important impact on guiding the exploration of unconventional oil and gas reservoirs.However,the acquisition condition for micro-seismic data is much complicated and with serious ambient noises,such as,environmental noise,mechanical vibrations,which results in weak energy of micro-seismic events.The signal-to-noise ratio(SNR)of the effective signal is very low and it is often submerged in the noise so that can't be recognized effectively.And the low SNR of the micro-seismic data limits the development of micro-seismic monitoring technology greatly.So,it is important to find new methods that can improve SNR of micro-seismic data effectively.In this paper,I present a method based on phase space reconstruction to process the real field micro-seismic data and achieve the purpose of denoising.By embedding the micro-seismic data into a suitable phase space and according to differences of geometric characteristics between effective signal and random noise in the phase space,the efficient micro-seismic signals can be recovered and the SNR of micro-seismic data can be improved.Then parameters of reconstruction have an important impact on the result of phase space reconstruction.So,I also study how to choose the embedding dimension,time delay and how to determine the neighbor radius.The C-C method is adopted to calculate the parameters of reconstruction and I proposed an iterative method to compute the radius.By adopting these methods,we can obtain much better results of reconstruction and denoising.In addition,I also study the automatic picking method of micro-seismic events and uses AIC,STA/LTA,AIC-STA and fractal dimension to achieve picking micro-seismic events automatically.Finally,this study shows that the denoising method has obvious effect in improving the SNR of real field micro-seismic data,especially for the lower SNR data.After denoising,the effective signal can be easily recognized and by relevant picking methods a better and more accurate signal can be acquired.