Research On Random Noise Suppression In Microseismic Exploration Based On Complex Shearlet Transform

In recent years,the trend of oil and gas exploration and development in China is mainly to exploit low permeability reservoirs,and the level of exploitation is also related to the main development trend of oil and gas industry in the future.However,the traditional seismic exploration technology is not ideal for data processing with low signal-to-noise ratio,which brings great difficulties to oil and gas exploitation.The signal-to-noise ratio of microseismic data is very low,and the effective signal is seriously disturbed by noise,which makes it difficult to identify.Therefore,noise suppression of microseismic data is an important step in seismic data processing,and also a prerequisite for imaging and interpretation of underground structures.Seismic exploration experts at home and abroad have carried out in-depth research on the suppression of random noise in microseismic data,and made some progress.However,the signal-to-noise ratio of ground microseismic data is very low,so the traditional filtering method has the problem of signal distortion when suppressing noise.Shearlet transform is a new multi-scale geometric analysis method.Its multi-scale and good directionality ensure the sparse representation of seismic signals and better recovery of seismic exploration signals when suppressing random noise.However,the redundant algorithm affects the computational efficiency of a large number of microseismic data processing.This paper presents a random noise suppression method for microseismic exploration based on complex Shearlet transform.Complex Shearlet transform combines the advantages of translation invariance and computational efficiency of dual-tree complex wavelet transform.In scale decomposition,dual-tree complex wavelet decomposition is used,and then shear filter is used to decompose the decomposed coefficients.Complex Shearlet transform not only guarantees translation invariance,but also speeds up the calculation speed of Shearlet transform.It is suitable for the analysis of microseismic exploration data and noise suppression.To solve the problem of low signal-to-noise ratio in microseismic exploration,a bivariate shrinkage denoising method based on quantum derivation theory is proposed in this paper on the basis of complex Shearlet transform.By analyzing the coefficients characteristics of complex Shearlet transform decomposition,it is found that the coefficients of parent and offspring after complex Shearlet decomposition have intra-layer correlation and inter-layer correlation.Then,a bivariate shrinkage denoising model is established by using the correlation between the coefficients of the parent and the offspring.According to the characteristics of microseismic random noise,combining with quantum derivation theory,a threshold is constructed according to the adaptive variation of noise characteristics.A bivariate shrinkage model of mixed denoising in high frequency,medium frequency and low frequency sub-bands is used to effectively recover the signal components at low signal-to-noise ratio in complex Shearlet domain.Theoretical analysis and simulation experiments show that the computation speed of complex Shearlet transform is three times faster than that of Shearlet transform.Meanwhile,the complex Shearlet transform can be used to represent microseismic signals more sparsely.The results of synthetic microseismic data and actual surface microseismic data processing show that the proposed algorithm can effectively suppress the random noise in microseismic data,restore the coherent phase axis more clearly,and improve the signal-to-noise ratio of microseismic exploration records.Compared with the traditional Shearlet transform bivariate shrinkage,the algorithm also has a remarkable improvement in the speed of operation,noise suppression and effective signal preservation,and has a good application prospect in the data processing of microseismic exploration.