小波神经网络在测井油水解释中的应用  

LOGGING INTERPRETATION OF OIL AND WATER WITH WAVELET-NN

作　　者：张学庆[1] 肖慈珣[1] 刘争平[2] 

机构地区：[1]成都理工大学信息工程学院,成都610059 [2]西南交通大学土木工程学院,成都610000

出　　处：《物探化探计算技术》2002年第2期115-118,191,共5页Computing Techniques For Geophysical and Geochemical Exploration

摘　　要：作者在本文中讨论了小波神经网络概念和油水测井曲线段的频率特性。以小波分析为数据预处理工具 ,首先用小波分解高频成分对测井曲线进行分段 ,然后应用小波分析提取出测井段的分频能量特征向量 ,以此向量代表井段的信息 ,并作为神经网络的输入 ,进行神经网络的测井解释 。This paper discusses the conception of waveletartificial neural networks and the frequency characters of well logging, analyze the logging data with wavelet analysis as preprocessing tool. First the logging curves are layered with their high frequency component abstracted by wavelet analysis; Then the characteristic vectors composed of energy of every frequency component are picked up and used as input vector of artificial neural networks for interpretation. With this method, we achieve the zoning oil-water interpretation using well logging data.

关 键 词：小波神经网络 小波分析 神经网络 测井解释 含油层 含水层 测井曲线 

分 类 号：P618.130.8[天文地球—矿床学] P631.8[天文地球—地质学]