基于神经网络的储层识别处理技术及在塔河油田的应用  

PROCESSING TECHNOLOGY OF RESERVOIR IDENTIFICATION BASED ON NEURAL NETWORK AND ITS APPLICATION IN THE TAHE OIL FIELD

作　　者：黄科[1] 肖慈珣[2] 张哨楠[1] 智慧文[1] 汪华[2] 

机构地区：[1]成都理工大学能源学院,成都610059 [2]成都理工大学信息工程学院,成都610059

出　　处：《物探化探计算技术》2006年第1期18-21,2,共4页Computing Techniques For Geophysical and Geochemical Exploration

摘　　要：简述神经网络用于储层识别模型的建立,并指出单独使用神经网络进行识别的不足。对此,设计了具有辅助神经网络作用的计算机程序,从而使单独使用网络的不足之处得到完善。通过塔河油田实际资料应用表明,在单独使用BP神经网络进行储层类型预测效果不是十分理想的情况下,采用这种处理技术,能够取得较好的效果。This article has summarized the establishment of the neural network model which can identify the reservoir type,and pointed out the limitation of using nerve network independently.In view of this,the computer program assisting the neural network has been designed,which can overcome the insufficiency of using neural network alone.It has been shown in practical application in the Tahe Oil Field that this kind of processing technology brings good prediction effect in an area in which BP network has not predicted with extremely ideal effect.

关 键 词：测井解释 计算机程序 神经网络 储层识别 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]