摘 要 HHT（Hilbert-Huang-Transform，希尔伯特-黄变换）是一种处理非线性、非平稳信号的时频分析方法，具有自适应性，能够解决以往傅立叶分析方法存在的一些问题，而且HHT分解所得到的固有模态函数分量具有明确的物理意义。根据其特性我们把它应用到地震资料的处理中。 本文的工作主要包括以下几个方面： （1）首先介绍几种常用的非平稳信号时频方法，然后阐述HHT的原理以及存在的一些问题，针对出现的端点效应问题，有目的地对算法进行改进。 （2）基于传统谱白化方法在提高地震分辨率上的不足，提出一种基于HHT的谱白化处理方法，理论模型数据的测试证明了该方法有效性和实用性，并将其运用到实际的地震资料中。 （3）在地震资料的时频分析时运用HHT方法，得到的Hilbert谱更加精细地刻画地震信号的形态特征。最后运用HHT方法提取地震属性剖面，得到IMF1分量的瞬时频率、瞬时振幅、瞬时相位剖面，并与原始的“三瞬”剖面进行对比。 本文的研究结果表明：通过HHT算法改进，可以有效抑制端点效应。在薄层或薄互层识别、微小构造的检测方面，基于HHT的谱白化方法能有效提高地震垂向分辨率；IMF1分量的瞬时频率剖面有助于对地下构造进行更好的解释。HHT有着很好的时频分析能力，在提高地震分辨率、提取属性方面显示了优越性，在地震资料处理解释领域发展空间较大，值得进一步探讨与研究。

ABSTRACT HHT(Hilbert-Huang Transform) is a way of time-frequency analysis methods, which can deal with nonlinear and the non-stationary signal.It is adaptive, fundamentally solves some existing problems based on Fourier analysis, and intrinsic mode function component obtained through decomposition also has specific physical meaning.According to its characteristics we put it into use in seismic data processing. The main work of the dissertation includes: (1)This paper introduces several conventional non-stationary signal time-frequency methods,then expounds the principle of HHT and some existing problems.According to the endpoint effect in algorithm, the algorithm is improved purposefully. (2)Based on the shortage of traditional spectrum methods in improving seismic resolution, a spectrum whitening method based on HHT is proposed.Theoretical model data is tested and the result shows the method is effective and practical.It will be applied to practical seismic data. (3)Using HHT in time-frequency analysis of seismic data,which produces the Hilbert spectrum that depicts more subtle to the morphological characteristics of the seismic signal.Finally extracting seismic attribute profile by using HHT method,which can get instantaneous frequency, instantaneous amplitude and instantaneous phase profile of imf component that compared with original “three instant” profile. In this paper the studies result show that the endpoint effect is effectively suppressed after the algorithm is purposefully improved.HHT is applied to the actual seismic data,which can effectively improve seismic vertical resolution in thin bed or thin interlayer bed identification and tiny structural detection aspect.Through comparing with the profile,the instantaneous frequency profiles of IMF1 component is helpful for better explaining the underground structure.HHT has better time-frequency analysis ability. It shows the superiority in improving seismic resolution and extracting three instantaneous seismic attribute.The prospect of HHT in seismic data processing and interpretation field is great,it is worth further discussing and researching.