高分辨时频分析方法在储层预测中的应用

发布时间：2018-03-04 08:26

本文选题：反褶积　切入点：短时傅立叶变换谱　出处：《成都理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：由于傅立叶变换是反映全局信息的变换,因此,无法表现信号的时间—频率的局部的特性。地震波在地下传播时,由于受到散射、吸收等影响,因而地震信号为非平稳信号。其统计量是随时间变化的函数,其频率特性也是改变的,一个单一的时间域或频域分析远不能满足实际的应用的需要。常规傅立叶变换只适用于平稳信号分析,不适用于非平稳信号,因而无法描述地震信号的时间—频率信息。因此,应该使用联合的信号时频分析特性分布的方法来展现地震信号的特征。近年来,时频分析已经是信号处理中的一个热点。采用时频分析方法,将时域转换到频域,可以在时频分布图显示出信号频谱与时间的关联特性。本文具体介绍了短时傅立叶变换、小波变换、S变换、广义S变换、基于匹配追踪的时频分析方法和反褶积短时傅立叶变换的基本原理以及优缺点。文中对几种理论信号进行多种时频分析方法的计算机仿真效果对比,结果证明,反褶积短时傅立叶变换与传统的时频分析方法相比更具优势。最近几年,利用地震信号的低频信息进行储层预测及油气识别方面取得新的进展,低频信息的应用越来越受到重视。理论及实践表明:地震波穿过流体时,地震信号的高频部分能量被吸收,而顶界面高频能量不会发生衰减,使得在单频剖面上含油气储层下方能量与含油气储层中上部分的能量出现差异。这个现象为低频阴影现象。在一定条件下,低频阴影可作为含油气性的一个直接指标。油气藏识别的有利标志为:低频时,上强下强;高频时,上强下弱。然而对于薄储层,低频阴影现象仍然较弱,因此有必要采用具有更高时频分辨率的时频分析方法进行低频阴影的识别。本文对三层水平层状模型进行低频阴影现象的数值模拟,由结果可知,反褶积短时傅立叶变换方法比广义S变换在薄储层预测中取得了更好的效果。笔者采用反褶积短时傅立叶变换和广义S变换两种时频分析方法对实际地震数据提取单频剖面,结合原始全频带地震数据分析进行油气检测。对比发现,高分辨时频分析方法(如反褶积短时傅立叶变换)取得更好效果。
[Abstract]:Because Fourier transform is a transformation that reflects global information, it can not express the local characteristics of time-frequency of the signal. The seismic wave is affected by scattering, absorption and so on when it propagates underground. Therefore, the seismic signal is a non-stationary signal. Its statistic is a function of time variation, and its frequency characteristic is also changed. A single time-domain or frequency-domain analysis is far from meeting the needs of practical applications. The conventional Fourier transform is only suitable for stationary signal analysis, not for non-stationary signals, so it is impossible to describe the time-frequency information of seismic signals. In recent years, time-frequency analysis has become a hot spot in signal processing. The time-frequency analysis method is used to convert time domain to frequency domain. The correlation between signal spectrum and time can be shown in time-frequency distribution diagram. In this paper, the short time Fourier transform, wavelet transform S transform, generalized S transform, generalized S transform are introduced in detail. The basic principle and advantages and disadvantages of time-frequency analysis method based on matching tracing and deconvolution short-time Fourier transform are presented. The computer simulation results of several theoretical signals are compared and the results show that, The short-time Fourier transform of deconvolution is superior to the traditional time-frequency analysis method. In recent years, new progress has been made in reservoir prediction and oil and gas identification by using low-frequency information of seismic signals. The application of low frequency information has been paid more and more attention. The theory and practice show that the high frequency partial energy of the seismic signal is absorbed when the seismic wave passes through the fluid, but the high frequency energy of the top interface does not decay. The difference between the energy under the oil and gas reservoir and the energy in the upper part of the oil and gas reservoir on the single frequency profile is shown. This phenomenon is a low frequency shadow phenomenon. Under certain conditions, Low frequency shadow can be used as a direct index of oil and gas bearing property. The favorable marks of oil and gas reservoir identification are: low frequency, strong and strong, high frequency, strong and weak, but for thin reservoir, low frequency shadow is still weak. Therefore, it is necessary to use time-frequency analysis method with higher time-frequency resolution to identify low-frequency shadows. The method of deconvolution short-time Fourier transform is more effective than generalized S-transform in thin reservoir prediction. Two time-frequency analysis methods of deconvolution short-time Fourier transform and generalized S-transform are used to extract the single frequency section from the actual seismic data. Combined with the original full-band seismic data analysis for oil and gas detection, it is found that the high-resolution time-frequency analysis method (such as deconvolution short-time Fourier transform) has better results.
【学位授予单位】：成都理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.4;P618.13

【参考文献】