Research Of Noise Suppression Method On Ground Microseismic Data

With the rapid development of economy,the demand for oil is increasing.However,at present,the exploitation potential of conventional oil and gas reservoirs is constantly decreasing,so the exploration and exploitation of unconventional oil and gas reservoirs has become an inevitable trend.Hydraulic fracturing technology is one of the important means to increase oil production and exploit unconventional oil and gas reservoirs.The microseismic monitoring technology is used to study the microseismic problems induced by hydraulic fracturing.But,the microseismic signals induced by force fracturing have different energy levels,weak signals,high frequencies,and short durations,which make the collected microseismic data have a low signal-to-noise ratio.In particular,it has a greater impact on the microseismic data obtained by selecting ground monitoring methods,and it is more severely affected by background noise.Too low signal-to-noise ratio will have a great impact on the accuracy and reliability of microseismic research and subsequent interpretation work.Therefore,solving the problem of noise suppression of microseismic data is a key step in microseismic data processing,and has important research significance for microseismic monitoring technology.Firstly,this paper describes the microseismic monitoring technology,and introduces two monitoring methods: ground monitoring and in-well monitoring.The characteristics of ground microseismic data and its main noise types are analyzed.The microseismic data can be synthesized using the ricker wavelet for algorithm simulation.The basic principles of common ground microseismic data noise suppression methods and their advantages and disadvantages are summarized.Based on the difference between the effective signals and noise interference of ground microseismic data,the microseismic data noise suppression methods are divided into five categories.Secondly,a time-frequency peak noise suppression method based on wavelet transform is improved.This algorithm aims at solving the problem of partial effective signal loss when the traditional time-frequency peak filtering method is used to suppress the noise of microseismic data,combining with the non-stationary characteristics of ground microseismic signals that can be effectively monitored by wavelet transform,a time-frequency peak filtering method based on wavelet transform is proposed.According to the main characteristics of ground microseismic data,a wide range of different wavelet basis functions and decomposition scales are selected to decompose the original noise data,and then combined with the time-frequency peak filtering method to suppress the random noise.The results show that the time-frequency peak filtering method based on wavelet transform can improve the signal-to-noise ratio of the microseismic data,improve the lineups in the image,enhance the resolution of valid signals,reduce the loss of effective signal and protect the edge characteristics of the signal when the wavelet base sym6 is selected and the scale three layers are decomposed.Thirdly,the singular value decomposition noise suppression method based on CEEMD(Complementary Ensemble Empirical Mode Decomposition)is improved.A new singular value decomposition filtering method based on CEEMD is proposed based on the advantages of complementary advantages and combining the respective characteristics of the two algorithms of CEEMD and singular value decomposition.First,CEEMD decomposition is performed on the microseismic data,and the obtained intrinsic mode functions are transformed into a Hankel matrix.Then,the median and mean values of singular values are selected as threshold values,and the singular value decomposition is performed on them.Finally,the signal is reconstructed to achieve noise suppression.The simulation experiments using the microseismic data synthesized by ricker wavelets show that using this method can effectively improve the microseismic image's lineups,suppress random noise in original noisy data,preserve the dynamics characteristics of the effective part of the microseismic signal,and reduce the loss of the effective signal and the distortion degree,improve the signal-to-noise ratio of data.Finally,Two improved algorithms for noise suppression in real ground microseismic data.The actual ground microseismic data is read by MATLAB software programming,and then the proposed two new noise suppression methods are applied to the single-channel microseismic data and multi-channel microseismic data processing.The effectiveness and applicability of the improved two noise suppression algorithms are verified,and the noise suppression of ground microseismic data is realized,the signal-to-noise ratio of the original data is effectively increased,and the lineups is improved.In addition,the two algorithms are used to process the same actual ground microseismic data,and the noise suppression results are compared and analyzed.Time-frequency peak noise suppression method based on wavelet transform is used to denoise the data with a higher signal-to-noise ratio,which can be better improving the lineups.A small amount of noise still exists in the low-frequency part after applying the singular value decomposition noise suppression method based on CEEMD,and the calculation time is long,but it can better protect the low-frequency part of effective signal and reduce the loss of effective signal.