摘要
基于灰色GM(O,N)静态模型和BP神经网络模型,提出了一种新的参数估算的模型即灰色神经网络模型来预测低渗砂岩储层的水锁损害。结果表明,用灰色神经网络模型来预测低渗砂岩储层的水锁损害是前行的。经验证,该预测模型具有较高的准确率,实用性强。该模型充分利用了灰色GM(0,N)静态预测模型建模弱化数据的随机性和累加数据的规律性以及神经网络的高度非线性,其预测效果明显优于传统的回归分析、灰色GM(0,N)静态预测法和一般的神经网络法。
Based on the gray GM(O,N)static model and BP neural network model,a gray neural network model used in predicting the formation damage of water blocking in low permeability sandstone reservoirs is established. By proof, the gray neutral network model used in predicting water blocking damage of low permeability sandstone reservoirs is reliable and has higher accuracy and practicability.The predicting result is better than the gray GM(O, N)model and the common neural network model.
出处
《钻采工艺》
CAS
2001年第1期38-40,共3页
Drilling & Production Technology
关键词
低渗储层
水锁效应
储层损害
灰色理论
神经网络
低渗透油田
low permeability reservoir, water lock effect, reservoir damage, grey theory, nerve network