论文详情

基于BP神经网络进行裂缝识别研究

断块油气田

2007年 14卷第02期

阅读：105

查看详情

Title

Research on frac ture identif ication based on BP
neural network

作者

申辉林高松洋

Authors

Shen Huilin, Gao Songyang.

单位

中国石油大学地球资源与信息学院, 山东东营257061

Organization

College of Georesources and
Information, China University of Petro leum, Dongy ing
257061, China

摘要

裂缝系统是个复杂的地质体, 其储层物性的改善作用是非线形的, 各种评价参数与裂缝发育程度之间的关系也是非线形的。基于人工神经网络理论, 开展了常规测井资料识别评价裂缝的研究。结果表明, 基于BP神经网络的裂缝性储集层常规测井识别, 与成像测井对比具有较好的应用效果。

Abstract

Fracture system is a comp licated geo log ic body. The
im provement e ffect o f reservo ir physica l property is nonlinear
and the re la tionsh ip be tw een the various eva luated param eters
and the extent o f fractures grow ing is nonlinear too. There are
the subjective uncerta inty and am bigu ity factors fo r using the
routine logg ing data to recogn ize fractures. A lthough the
im ag ing logg ing are intu itive and accurate, the cost are very
h igh. The resu lt show s tha t fractures iden tify ing based on BP
neura l ne tw ork has preferab le effect com pared w ith im ag ing
logg ing.

关键词：

裂缝识别常规测井成像测井人工神经网络;

Keywords:

fracture identification, routine logg ing,
im ag ing logg ing, artific ia l neu ra l netwo rk.;

基金项目

DOI