基于CEEMD的地震数据小波阈值去噪方法研究

考虑到地震信号的非平稳性和去噪方法对非平稳信号的适应性,针对互补集合经验模态分解(CEEMD)舍弃高频分量的去噪方法和小波阈值去噪方法存在的不足,提出了基于CEEMD的地震数据小波阈值去噪方法。CEEMD是EMD(经验模态分解)的改进型算法,它既保留了EMD处理非平稳信号的优势,又能有效地克服EMD的模态混叠问题;但是,单纯的CEEMD分解去噪会在去除高频噪声的同时压制高频的有效信息。将CEEMD分解与小波阈值去噪相结合,对CEEMD去噪要舍弃的含噪声较多的高频固有模态函数(IMF)分量进行小波阈值去噪,以保留这些分量中的有效信息。模型数据和实际地震资料的测试结果表明,无论对于低噪声还是强噪声地震数据,基于CEEMD的小波阈值去噪方法的去噪效果都优于单纯的CEEMD去噪方法和小波阈值去噪方法。

Considering the non-stationary of seismic signal and the adaptability of denoising method to non-stationary signal,the wavelet threshold denoising method based on CEEMD is proposed to make up for the disadvantages of Ensemble Empirical Mode Decomposition(CEEMD) denoising method and wavelet threshold denoising method.CEEMD is an improvement algorithm of the Empirical Mode Decomposition(EMD).CEEMD has retained the advantages of the EMD in processing non-stationary signal and overcome the modal aliasing problem of EMD effectively.But the pure CEEMD denoising could suppress effective high-frequency information while removing high frequency noise.CEEMD is combined with wavelet threshold denoising method and wavelet threshold method is used to denoise the Intrinsic Mode Function(IMF) components with more noise that used to be abandoned for CEEMD denosing,in order to retain the effective information of the components.The test results of model data and real seismic data show the denoising effect of the wavelet threshold denoising method based on CEEMD is better than the pure CEEMD and wavelet threshold method whether for low-noise or strong noise seismic data.