Hilbert-Huang变换与傅里叶变换在大地电磁数据处理中的适用性分析  

Applicability analysis of Fourier transform and Hilbert-Huang transform in the procession of magnetotelluric signal

作　　者：蔡剑华[1] 肖永良 

机构地区：[1]湖南文理学院洞庭湖生态经济区建设与发展湖南省协同创新中心,常德415000 [2]湖南财政经济学院信息技术与管理学院,长沙410205

出　　处：《地球物理学进展》2017年第4期1548-1555,共8页Progress in Geophysics

基　　金：国家自然科学基金项目(41304098);湖南省自然科学基金项目(2017JJ2192;2017JJ2015);湖南省教育厅重点项目(16A146)联合资助

摘　　要：Hilbert-Huang变换(HHT)和傅立叶变换是目前广发应用于大地电磁(MT)信号处理的两种算法,但两种方法在MT信号的处理中的适用性研究却鲜有报道.文章以仿真平稳信号、加噪信号、非平稳信号和实测大地电磁信号为例,从准确性、稳定性、计算效率等几个方面比较了两种算法在大地电磁信号处理中的适用性.结果表明:傅立叶变换对于无噪平稳信号的分析,其分辨率和准确性很高,且计算速度快,适合海量大地电磁测深数据的处理;HHT算法具有时频分析和滤除高频分量的能力,能精确的刻画信号能量随时间和频率的分布,且抗噪声性能好,在MT数据筛选和去噪等方面有优势;基于HHT边际谱的功率谱估计更适合MT信号非平稳特性的实质,但其计算效率低,是制约其工程应用的瓶颈问题.Fourier transform and Hilbert-Huang Transform（HHT）are now widely used in magnetotelluric（MT） as two kinds of signal processing algorithm,while the applicability of the two methods is seldom reported.Taking the simulated stationary signal,noised signal,non-stationary signal and the measured MT signal as object of study,the applicability of two kinds of algorithm are compared from the accuracy,stability,computational efficiency and so on.The results show that for stationary signal without noise,the resolution and the accuracy of Fourier transform is high,and the calculation speed is fast.Fourier transform is suitable for mass MT sounding data processing.HHT algorithm has the ability of timefrequency analysis and filtering the high frequency component.HHT spectrum can describe the distribution of signal energy with time and frequency.Compared with Fourier method,HHT has the superiority in the MT filtering and data de-noising.Power spectrum estimation method based on HHT marginal spectrum is more suitable with the non-stationary characteristics of MT signal,but its computational efficiency is low and it is a bottleneck which restricts its engineering application.

关 键 词：傅立叶变换 HILBERT-HUANG变换 功率谱估计 时频分析 

分 类 号：P631[天文地球—地质矿产勘探]