摘要
利用BP人工神经网络误差反向传播算法 ,开发了神经网络油气层解释软件。通过对泌阳凹陷安棚深层系和焉耆盆地已试油层的原始资料进行学习、训练 ,建立了油气层神经网络解释模型。运用该模型可完成储层流体类型的划分和识别 ,结合录井、测井等原始资料 ,可实现计算机处理自动化 ,其预测符合率达 84 .2 %。
The logging nerve network interpretation software for oilgasbearing formations is developed using error backpropagation algorithm of artificial BP nerve network. The reservoir interpretation model with nerve network is built through studying and training of initial data of tested oilbearing formations in deep Anpeng series of strata in Biyang sag and Yanqi basin. This model can divide and recognize the type of reservoir fluid. It makes the computer automatically process the data combined with the original logging data. Its coincidence rate is up to 84.2%.
出处
《油气地质与采收率》
CAS
CSCD
2003年第2期36-37,79,共3页
Petroleum Geology and Recovery Efficiency
关键词
录井
神经网络
油气层
解释模型
研究
artificial BP nerve network, training, interpretation model, recognition of oilgasbearing formation
