石油学报 ›› 2001, Vol. 22 ›› Issue (6): 34-36.DOI: 10.7623/syxb200106007

• 地质勘探 • 上一篇    下一篇

利用地震属性预测渗透率

何琰1, 彭文2, 殷军2   

  1. 1. 西南石油学院勘探系 四川南充637001;
    2. 四川石油地调处 四川南充637001
  • 收稿日期:2001-04-05 修回日期:2001-07-10 出版日期:2001-11-25 发布日期:2010-05-21
  • 作者简介:何琰,女,1971年5月生,2000年毕业于西南石油学院,获博士学位,现为西南石油学院讲师.
  • 基金资助:
    国家“863”项目(863-306-04-03-1A)和“油气藏地质及开发工程国家重点实验室”开放基金项目部分研究成果.

Permeability prediction by seismic attribute data

HE Yan1   

  1. Southwest Petroleum lnstitute, Nanchong 637001, China
  • Received:2001-04-05 Revised:2001-07-10 Online:2001-11-25 Published:2010-05-21

摘要: 储层岩石的渗透率分布的不均匀性直接影响油气分布、运移以及开采。目前,在井资料十分有限的勘探初期,利用地震资料难以预测渗透率。因此,探讨利用地震属性预测渗透率具有重要的理论和实际意义。地震属性数据预测渗透率的Rough Set(RS)与Genetic Algorithm BP(GABP)(粗集与遗传BP神经网络)相结合的方法预测渗透率得到了令人满意的结果。表明:1利用地震属性在勘探早期预测储层的渗透率是完全可能的;2应用实例表明,该方法的预测精度较高。因此,本方法将成为地震数据预测渗透率的有效手段。

关键词: 渗透率, 地震属性, 属性优化, 遗传BP神经网络

Abstract: The heterogeneity of the permeability distribute of rock directly influences the distribution,migration and production of hydrocarbons.At present,it is hardly possible to apply seismic data to predict permeability in the early period of exploration when the well is not enough for founding the statistical relationship. In this paper,it is very significance to discuss the problem that apply the seismic attribute data to predict permeability.We briefly introduce the procedure of the RS and GABP to predict permeability.The application result indicates that:1.it is possible to apply seismic attribute data to predict permeability in exploration earlystage; 2.the method is very effective.So it can become an effective means for permeability prediction with seismic data.

Key words: permeability, seismic attribute, attribute optimization, genetic BP neural network

中图分类号: