论文详情
泥页岩分类的BP神经网络方法
石油钻采工艺
2003年 25卷 第4期
阅读：73
查看详情
Title
PREDICTION OF THE SHALE PERFORMANCE BY THE BP NEURAL NETWORK
Authors
Liu Bo
Yan Jienian
Li Jing
单位
石油大学石油天然气工程学院, 北京 102249
中国石化石油勘探开发研究院, 北京 100083
摘要
为探讨X衍射资料与泥页岩理化性能之间的定性关系,建立一种智能化模型对泥页岩的类型进行快速识别。在分析泥页岩理化性能与X射线衍射资料数据关系的基础上,提出了用BP神经网络预测泥页岩理化性能参数的方法及相应的模型。并针对BP神经网络算法进行了改进,使模型网络训练时的收敛速度比常规方法快了4倍以上。通过现场试验数据验证,该模型预测符合率较高,能够满足实际应用的要求。
Abstract
To analyse the relation between the X-Ray data and the shale performance.The model of BP neural network was designed here to predict the shale performance from by the X-Ray data.This model has been modified for good constringency rate more 4 times than that of the conventional methods.Thought the validation of the oil fields data,the model has a good performance and can meet the demand of the practical application.
关键词：
神经网络;
泥页岩;
预测;
理化性能;
井壁稳定;
模式识别;
Keywords:
neural network;
shale;
prediction;
shale performance;
well stability;
pattern recognition;
DOI
10.3969/j.issn.1000-7393.2003.04.007
论文详情