基于希尔伯特-黄变换的地震信号时频谱分析

Time-frequency spectral analysis of seismic data based on Hilbert-Huang transform

希尔伯特-黄变换(Hilbert-Huang Transform,HHT)是分析非线性、非稳定信号的一种新方法,能清晰地刻画地震信号的时频能量分布。首先将信号分解为有限数量的固有模态函数I MF,再对这些I MF求解瞬时频率,进而获得信号的时频谱。应用理论模型和实际地震道数据进行了试算,并与S变换谱进行了对比,证明该方法比S变换具有更好的时频域刻画能力。对实际二维地震剖面做HHT变换求得希尔伯特谱,提取分频剖面分析认为,HHT瞬时谱具有一定的油气检测能力。

Seismic data are nonlinear and nonstationary signals.Hilbert-Huang transform is a new method to analyze nonstationary signals.Signals are decomposed into a limited number of intrinsic mode functions.From these IMFs,instantaneous frequency was calculated,and the time-frequency spectra of the signals were finally determined.We verified Hilbert-Huang transform analysis on synthetic and real seismic data,and compared its results with those from S-transform analysis.It is showed that Hilbert-Huang transform is superior to S-transform in revealing the characteristics of time-frequency spectrum,and the instantaneous spectrum of Hilbert-Huang transform is a good indicator of hydrocarbon accumulation.