频谱重构技术及其在储层预测中的应用

发布时间：2018-06-16 02:22

本文选题：最小二乘反演 + 共轭梯度　；参考：《中国石油大学(华东)》2015年硕士论文

【摘要】：油气田勘探开发的程度越来越高,并且勘探领域不断扩展,所以难度逐渐增强,为了能够为油田的滚动勘探开发提供丰富有效的资料,我们提取出地震资料中更多的隐藏信息,地震属性分析能够满足这个要求。基于最小二乘反演的频谱重构技术假设地震资料频谱由雷克子波分量频谱线性组成,本文通过最小二乘反演迭代拟合,结合共轭梯度法求解非线性方程组,重构出雷克子波分量的主频值与振幅值从而能够求得不同主频雷克子波分量。针对不同地质构造对不同的频率特性敏感性不同,将频谱重构结果进行两项应用。一是对三维数据沿层提取数据,每道进行频谱重构,将求得主频值与振幅值作为两种新的属性,频谱重构主频属性和频谱重构振幅属性,应用到地震属性分析。利用这两种属性沿层进行了聚类分析,并且得到了良好的聚类效果,并同其他的属性效果进行了对比。二是对二维剖面求取平均频谱,并求取平均频谱的雷克子波分量,将得到的不同主频的雷克子波分量作为滤波器进行带通滤波,滤波后深层构造更加清晰。为了探索新提取的两种属性之间的联系,利用了主成分分析方法进行地震属性的优化。
[Abstract]:The degree of exploration and development of oil and gas fields is getting higher and higher, and the exploration field is expanding, so the difficulty is gradually increased. In order to provide rich and effective data for rolling exploration and development of oil fields, we extract more hidden information from seismic data. Seismic attribute analysis can meet this requirement. The spectrum reconstruction technique based on least square inversion assumes that the spectrum of seismic data is composed of linear components of Rayleigh wavelet components. In this paper, the nonlinear equations are solved by iterative fitting by least square inversion and conjugate gradient method. The main frequency value and amplitude value of the main frequency component can be obtained by reconstructing the main frequency wavelet component. Because different geological structures have different sensitivity to different frequency characteristics, the spectrum reconstruction results are applied in two applications. The first is to extract the data along the layer of 3D data and reconstruct the spectrum of each channel. The main frequency value and amplitude value are taken as two new attributes, the main frequency attribute of spectrum reconstruction and the amplitude attribute of spectrum reconstruction are applied to seismic attribute analysis. The two attributes are used to cluster analysis along the layer, and good clustering results are obtained, and compared with other attributes. The second is to obtain the average frequency spectrum of the two-dimensional section, and to obtain the components of the average spectrum. The different main frequency components of the Recker wavelet are used as the filter for bandpass filtering, and the deep structure of the filter becomes clearer. In order to explore the relationship between the two newly extracted attributes, principal component analysis (PCA) is used to optimize seismic attributes.
【学位授予单位】：中国石油大学(华东)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.4;P618.13

【参考文献】