摘要
将人工神经网络模型引入综合录井油气层评价中,提出了一种改进的BP算法。采用层次模式识别的方法,对塔里木油田的实际数据进行了试算,以识别油层、气层、含水油层、油水同层、含油水层、气水同层、油气水同层、水层、干层。实验结果表明:该方法可行,且不受气测参数有效范围的限制。
An artificial neural network model is introduced to oil and gas evaluation in comprehensive logging,and a kind of improved BP algorithm is concluded for advancing oil and gas evaluation and enhancing the coincidence rate of oil and gas interpretation By using a stratified pattern recognitionmethod, some experiments to recognize various kinds of oil and gas formations in Tarim oilfield are made The result shows that this method is feasible,and the recognition is not bound to the limited range of gas parameters
出处
《江汉石油学院学报》
CSCD
北大核心
1997年第4期112-114,共3页
Journal of Jianghan Petroleum Institute
关键词
录井
油气层
人工神经网络
油气勘探
gas logging
logging
reaervoir resarch
neural networks
pattern recognition V 1=V 2=2400 bps V 1=V 2=9600 bps aH b