基于Haar小波变换的渐弃型废弃河道相识别

Haar wavelet transformation-based gradually abandoned channel facies identification

针对测井曲线进行沉积微相分析和渐弃型废弃河道微相计算机自动识别的问题,提出基于Haar小波变换的测井曲线沉积微相和渐弃型废弃河道识别的方法.在该方法中,应用小波分解将测井曲线的信号分解为宏观特征信息、细节信息和噪音信息三部分;将渐弃型废弃河道和普通河道作为同一类,利用Bayes判别分析对经小波变换后得到的SP、RMN、RMG曲线的宏观特征信息建立其判相模型,通过该模型将渐弃型废弃河道和普通河道与其他沉积微相进行分离;利用渐弃型废弃河道和普通河道沉积微相在微电极幅度差上的差异,将RMN–RMG值与地层深度建立拟合曲线,利用曲线的斜率为判相依据,识别出渐弃型废弃河道和普通河道.采用C/C++语言在C++Builder平台上编写了本方法的应用程序,实现了从测井曲线沉积微相分析到平面相图自动生成的计算机化.该方法已被应用于大庆油田南二区东块1 480口井和南一区20口井的测井曲线分析处理,其结果与人工识别成果对比表明准确率达80%以上,能满足油田实际生产的要求.

This work exploited the application of Haar wavelet approach to sedimentary microfacies analysis using log data and gradually abandoned channel facies identifying.This method first,uses wavelet transformation to separate noise and minutiae from large scale patterns for log data.Then normal channels and gradually abandoned channels are regarding as the same,distinguished from the other sedimentary microfacies by Bayesian model created by SP,RMN and RMG values obtained from the large scale patterns.Finally,according to difference in the RMN-GMG between the two microfacies,the slope of the best fit line of RMN-RMG against depth is used to identify gradually abandoned channels and normal ones.A program to implement the method was developed in C/C＋＋ on C＋＋ Builder platform,which computerizes the sedimentary microfacies analysis with log data and facies ichnography creation.This highly computerized,easy-operating method has been used to analyze the log data of 1 480 wells in the east part of No.2 south area and 20 wells in No.1 south area in Daqing Oil Field,and the result was well consistent with that of manual work.The success rate of the procedure in correctly identifying the sedimentary microfacies is 80%,which shows its great potential for use in the exploration of meandering depositional oil fields.