人工神经网络油气预测技术在四川气区的应用  

18 Application of Artificial Nerve Network for Oil-gas Prediction in Sichuan Gas Fields

作　　者：巫盛洪[1] 高如曾[1] 李志荣[1] 韩德贵 

机构地区：[1]四川石油管理局地质调查处研究所

出　　处：《天然气工业》1996年第4期18-21,共4页Natural Gas Industry

摘　　要：介绍了根据人工神经网络模型的原理，将地震道信息引入人工神经网络系统进行油气预测的方法。首先从地震时窗内提取出反映储层油气变化的５组共１０多个地震特征参数，然后进行ＢＰ神经网络的自学习，进而结合地质情况和实钻资料进行油气水预测判别，最后将成果绘成图件。将该方法分别应用于四川气区的川南、川中和川东的３个试验区块。Based on the principle of artificial nerve network model a new way of oil-gas pradiction by inaroducing the trace information into the artificial nerve metwork is proposed out.first,extract 5groups.,including more than 10 seismic characteristic parameters that reflect the oil-gas variation of reser-voir from seismic time window;then fulfill self-study of the BP nerve network and carry out water-oil-gasprediction by the combination of geoloyic characteristics and real drilling data;finally,draw the diagram.Application of this method in Sichuan gas fields gets good results.

关 键 词：人工智能 神经网络 油气分布 储集层 四川 气田 

分 类 号：TE37[石油与天然气工程—油气田开发工程] P618.130.2[天文地球—矿床学]