经验模态分解法在地震资料处理中的应用试验

An empirical mode decomposition method in seismic data processing of application and testing

通过对多种连续小波基的平稳序列时频分析,提出经验模态的非平稳序列的时频分析方法,并在地震资料处理解释应用中作探索性试验。试验结果表明:对地震记录道长为5 s和2 ms采样资料,经验模态分解后可获得6—7个固有模态函数单分量,分解的第一个分量C1(t)具有较宽的频率成分,可作为地震属性的主分量。

In this article, based on the continual wavelet base time-frequency analysis method of stationary series, an time-frequency analysis method of non-stationary series with empirical mode decomposition is presented. And an interpretation of seismic data processing applications was tentatively tested. The results show that from a seismic trace of 5 seconds length and 2 ms sampling,6--7 IMF single components can be obtained after decomposed, the first component C1 （t） of the decomposition, which is with wide frequency components, can be used as the principal component analysis of seismic data.