西南石油大学学报(自然科学版) ›› 2002, Vol. 24 ›› Issue (3): 28-30.DOI: 10.3863/j.issn.1000-2634.2002.03.008

• 地质勘探 • 上一篇    下一篇

人工神经网络在录井油气水层识别中的应用

钟大康1 张国喜2
  

  1. (1.石油大学资源与信息学院,北京 昌平 102249; 2.中国石油录井公司,北京 100043)
  • 收稿日期:2001-09-23 修回日期:1900-01-01 出版日期:2002-06-30 发布日期:2002-06-30
  • 通讯作者: 钟大康

APPLICATION OF ARTIFICIAL NEURAL NETWORK ON THE OIL-GAS-WATER RECOGANITION IN THE WELL LOGGING

ZHONG Da-kang1 ZHANG Guo-xi2 . 
  

  1. (University of Petroleum, Changping Beijing 102249,China)
  • Received:2001-09-23 Revised:1900-01-01 Online:2002-06-30 Published:2002-06-30
  • Contact: ZHONG Da-kang

摘要: 地质录井油气水层识别有多种方法,如岩心岩屑肉眼观察法、录井数据图版法、线性函数判别法等。主要介绍人工神经网络法,内容包括方法的基本原理、系统结构组成、判别时选用的参数、应用效果与结果分析。将该方法用于准噶尔盆地不同地区进行实际检验,结果表明:该方法对未知样品的正判率均在90%以上,明显优于传统方法,取得了良好的应用效果,证明该方法在这一领域具有广阔的应用前景。

关键词: 神经网络, 油气水层识别, 地质录井, 准噶尔盆地

Abstract: There are many kinds of oil-gas-water recognition methods such as the core and cuttings testing, cross charting, linear function recognition. This paper mainly introduces the artificial neural network methods, which include the basic principal,composition of the system, recognition parameters, result and
effectiveness of the methods. It has been applied to a few oil fields in Jungar basin. The result indicates that the recognition precision is over 90%, apparently superior to the traditional methods, and the methods have a broad application prospect in this area.

Key words: neural network, oil-gas-water discrimination beds identification, geological logging, Jungar basin

中图分类号: