Research On Reconstruction Method Of Irregular Seismic Data Based On Compressed Sensing

In recent years,with the deepening and development of oil and gas exploration in China,the objects of seismic exploration have become more and more complex,which puts higher requirements on the processing quality of seismic data.Due to the influence of complex geological conditions during the acquisition and processing of seismic data,the seismic data may be irregular and incomplete.The subsequent processing of seismic data has high requirements for the regularity and integrity of seismic data,so it is critical and necessary to regularize and reconstruct seismic data.For irregular seismic data,it can be divided into three types:1.Sparse seismic data;2.Regular seismic data with random missing data;3.Complex irregular seismic data.According to the performance characteristics of different types of irregular seismic data,a targeted reconstruction method can be studied.This paper mainly studies the data reconstruction methods for two kinds of irregular data,such as Regular seismic data with random missing data and Complex irregular seismic data.For the regular seismic data with random missing data,this paper builds a seismic data interpolation reconstruction method based on compressed sensing theory,combined with the current technology is widely used in the Fourier transform.By introducing the Bregman iterative algorithm and combining different threshold model parameters,the rapid reconstruction of irregular missing seismic data is realized.For complex irregular seismic data,we first regard the regularized reconstruction of irregular seismic data as an inversion problem,and combine the non-uniform Fourier transform with the least squares inversion method to realize the regularization of two-dimensional irregular data.For the non-uniform Fourier transform,the reconstruction result is not ideal when the curvature of the event is large.We propose an irregular seismic data reconstruction method based on the anti-leakage Fourier transform.Through the recursive phase decompression system spectrum leakage phenomenon,the Fourier basis function is re-orthogonal,and the relatively accurate Fourier coefficient is estimated,so as to achieve the purpose of regular reconstruction data.The results show that the reconstruction method based on anti-leakage Fourier transform has certain anti-aliasing ability,and the reconstruction ability of complex irregular seismic data is better than non-uniform Fourier transform reconstruction method.We used synthetic seismic records and actual work area data to confirm the effectiveness of the above method.