【作者】 薛志刚；

【导师】 乐友喜；

【作者基本信息】 中国石油大学 ， 地球探测与信息技术， 2010， 硕士

【摘要】 时频分析的基本思想就是设计时间和频率的联合函数,用其同时描述信号在不同时间及频率的能量分布情况,使得信号的局部特征能够在时间-频率二维平面内清晰地显示出来。本文首先从傅立叶变换入手,分析其优点及其局限性;然后结合实例研究了短时傅立叶变换、小波变换以及Wigner-Ville分布,它们都是比较经典的时频分析方法,但是在实际应用中还是存在一定的问题;研究了最近兴起的希尔伯特—黄变换方法,其通过经验模态分解将信号按一定的顺序分离出来,有很好的时间和频率分辨率,但是由于其自身原因存在端点飞翼现象,使得分离出来的信号不能很好地复原。本文重点讨论了S变换的方法原理、算法及其性质,通过引入两个可变的调节参数来改造其高斯时窗函数,得到了广义S变换,并通过模型阐述了调节参数对广义S变换时频分辨率的影响以及如何选取最佳的参数。模型计算结果表明,广义S变换具有很好的时频聚集性能,而且能够比较准确地显示出信号的局部特征,同时可以识别薄层模型的顶底,为实际地震数据的高分辨率处理提供了依据。在实际地震资料的应用中,应用高精度时频分析方法划分沉积旋回,通过调谐频率加强法实现了针对小尺度地质目标的地震资料高分辨率处理,处理后的资料应用于地震属性分析、地震相分析以及波阻抗反演,预测结果得到了一定程度的改善,从而提高了地震资料识别和描述储层的能力。更多还原

【Abstract】 The basic idea of time-frequency analysis is to design the joint function of time and frequency,which is used to describe the different times and frequency of the energy distribution that can make the local characteristics of the time-frequency two-dimensional plane clearly displayed.This thesis start from the Fourier transform to analyze the advantages and limitations;then with an example of a short-time Fourier transform, wavelet transform and Wigner-Ville distribution, they are more classic methods of time-frequency analysis, but there are some problems in practical application;the recent emergence of Hilbert-Huang transform method,which separate the signal by a certain order through empirical mode decomposition,have a good time and frequency resolution, but due to reasons of its own ends flying wing phenomenon that makes the separation of the signal can not properly recover.This thesis focuses on the method, the algorithm and the nature of S transform that is introduced to two variable parameters to regulate the Gaussian time window function,which is called generalized S transform, through the model the thesis expatiate the adjust parameters on the impact of time-frequency resolution and how to select the best parameters.Through model analysis, the generalized S transform has good time-frequency performance and can accurately show the local characteristics of the signal, while the model can identify the top and bottom of the thin, as the actual seismic data processing provides high-resolution basis.In real applications, the time-frequency analysis is used to divide sedimentary cycles and achieve by tuning the frequency to strengthen law on a small-scale geological targets high-resolution seismic data processing.The processed data is used in seismic attributes analysis, seismic facies analysis, and impedance inversion, the predictions have been improved to some extent, which improve the capacity of identification and description of reservoir.更多还原