BP神经网络在测井解释中的应用研究  

Research on the Application of BP Neural Network in Wells log Interpretation

作　　者：李映涛[1] 袁晓宇[1] 刘迪[1] 孟祥豪[1] 

机构地区：[1]成都理工大学油气藏地质及开发工程国家重点实验室,四川成都610059

出　　处：《西部探矿工程》2013年第3期45-48,共4页West-China Exploration Engineering

基　　金：油气藏地质及开发工程国家重点实验室基金(编号:PLC201201)资助

摘　　要：依据塔中各井的测井数据,采用BP神经网络算法对各井的孔隙度进行预测,并初步划分有利储集体分布范围。在实际应用中,根据研究区的测井曲线类型,采用了3种模型进行孔隙度预测实验,通过试验、分析,最终优选模型二对各井的孔隙度进行预测。分层位制作了孔隙度平面分布图,通过对孔隙度的分析可知,塔中地区良里塔格组及鹰山组地层普遍发育,但总体为低孔低渗致密碳酸盐岩,局部发育孔隙度较好储集体。这与岩芯分析结果一致,也证实了该方法的有效性。Based on the well log data in Tazhong area, the paper used BP neural network method to predict porosities of these wells, and preliminarily divided favorable reservoir distribution. In practical applications, according to the type of well logging curves, we adopted three kinds of model to make porosity predic- tion experiments. Through the test and analysis, the second model is ultimately chosen to explain porosity in wells. We made plane distribution maps of porosity in different layers, and the porosity analysis shows that, Lianglitage formation and Yins- hang formation develop widely in Tazhong area, but as a whole, they are carbonate rocks with low porosity and low permeability, locally is reservoirs with better porosity developed. It is consist- ent with the rock core analysis result, and demonstrates the ef- fectiveness of the method.

关 键 词：BP神经网络 测井孔隙度 有利储集体分布 

分 类 号：TE358[石油与天然气工程—油气田开发工程]