Study On The Separation Methods For Multi-source Blended Seismic Acquisition Data By Sparsity-promoting Constraints

The simultaneous-source technique aims to removing the limitation of no interference between adjacent shots by allowing more than one source to be shot simultaneously,thus greatly shortening the construction period,improving the spatial sampling,and improving the imaging quality of underground structures.At present,there are two main processing methods for the blended data,one is to directly apply migration processing without deblending,the other is to deblending firstly,and then apply conventional processing scheme.This article main research content for the inversion method based on sparsity-promoting constraints,first of all,deblending technique development present situation at home and abroad made a comb,blended data acquisition are analyzed,the basic principle of deblending using normal moveout and median filtering in common-midpoint gathers is given,a method of deblending using space-varying median filter is studied.The two median filtering effects are compared with model data.The results show that the space-varying median filter has better effect in processing the skew coaxial blended data.Then sparsity-promoting constraint iterative threshold shrinkage algorithm of deblending is discussed,we select FFT?Curvelet and Seislet three sparse transform to verify the effectiveness of the method described in this article by model data and field data.The results show that the Seislet transform has better sparsity than the FFT and Curvelet transform,and has better performance in processing blended data.In the end,we systematically studied parameters such as iteration cycles,threshold parameter selection and dip estimation for deblending effects using Seislet transform.