SVD adaptive seismic data noise suppression method

Abstract

The invention relates to an SVD adaptive seismic data noise suppression method, which comprises the steps of: firstly, carrying out smooth preprocessing on data by using an average filter, and markingstrong periodic noise channels according to the difference between average energy values of all channels; secondly, rounding an average time period mean value of each channel to serve as a column number of a single-channel constructed Hankel matrix, and enhancing the correlation between the channels of the reconstructed matrix; and finally, performing SVD on each reconstruction matrix one by one,inputting singular values into a constructed K-K joint network for identification and distinguishing so as to determine values corresponding to different correlation strength degrees in the signals,reserving valid signal singular values for data reconstruction, and completing adaptive noise suppression in the time domain. According to the SVD adaptive seismic data noise suppression method, the K-K joint network is introduced to determine the SVD reconstruction order, so that the situation of depending on human experience is reduced. The SVD adaptive seismic data noise suppression method cansuppress periodic noise and random noise which always penetrate through single-channel data such as power frequency interference and are high in regularity at the same time for shallow stratum seismicdata, so as to improve the seismic data quality.

Description

technical field [0001] The invention belongs to the field of denoising seismic data in shallow strata, and aims at strong periodic noise and random noise in seismic data, in particular an SVD adaptive seismic data noise suppression method based on K-K joint network. Background technique [0002] The shallow stratum seismic data collected in the field contains a large amount of information such as underground structure and lithology, as well as random noise, and strong periodic noise such as power frequency interference throughout the data, that is, there is a mixture of effective information and different noises. Stacking directly affects the quality and signal-to-noise ratio of seismic data, which is not conducive to further analysis and processing of data. SVD analysis technology, one of the data processing methods, has been applied to noise suppression of seismic data by many scholars because of its advantages such as easy implementation. The essence of SVD is an orthogo...

AI Technical Summary

Problems solved by technology

Article "Denoising method of surface microseismic data based on single-channel SVD and amplitude ratio" (Hu Yongquan, Huang Jianbo, Tian Zhihua, Pan Shulin. Denoising method of surface microseismic data based on single-channel SVD and amplitude ratio[J]. Petroleum Geophysical Prospecting, 2019,58(01):43-52+62.) Simultaneous suppression of strong periodic interference and random interference in surface microseismic data, pointing out that the large, medium, and small parts of the singular values ​​correspond to periods with strong correlation Signals, effective signals with a certain correlation, and random interference without correlation, the reconstruction singular value is compared with the reconstruction results of different selection ratios by trial and error, which reduces the data processing efficiency

Images