石油学报 ›› 2015, Vol. 36 ›› Issue (11): 1449-1456.DOI: 10.7623/syxb201511015

• 特约稿件 • 上一篇    下一篇

基于知识库的解释模型智能优选测井数据处理方法

尚福华1, 原野2, 王才志2, 曹茂俊1,2, 冯周2   

  1. 1. 东北石油大学计算机与信息技术学院 黑龙江大庆 163318;
    2. 中国石油勘探开发研究院 北京 100083
  • 收稿日期:2015-06-06 修回日期:2015-09-07 出版日期:2015-11-25 发布日期:2015-12-05
  • 通讯作者: 曹茂俊,男,1978年5月生,2001年获大庆石油学院计算机专业工学学士学位,现为中国石油勘探开发研究院地球探测与信息技术专业博士研究生,主要从事测井数据处理与解释、软件工程、知识库等相关研究。Email:caomaojun@126.com
  • 作者简介:尚福华,男,1962年12月生,1983年获大庆石油学院数学专业理学学士学位,2007年获哈尔滨工业大学计算机专业工学博士学位,现为东北石油大学计算机与信息技术学院教授,主要从事人工智能、软件工程、知识库、油田智能软件等相关研究。Email:shangfh@163.net
  • 基金资助:

    国家重大科技专项(2011ZX05020)、国家自然科学基金项目(No.61170132)和中国石油天然气集团公司重大科技专项(2013E-3809)资助。

A logging data processing method of intelligent optimization logging interpretation model based on knowledge-base

Shang Fuhua1, Yuan Ye2, Wang Caizhi2, Cao Maojun1,2, Feng Zhou2   

  1. 1. School of Computer & Information Technology, Northeast Petroleum University, Heilongjiang Daqing 163318, China;
    2. PetroChina Research Institute of Petroleum Exploration & Development, Beijing 100083, China
  • Received:2015-06-06 Revised:2015-09-07 Online:2015-11-25 Published:2015-12-05

摘要:

在复杂储层评价中,针对不同性质特点的目标层段自动选取最佳校正图版和解释方程,这对测井曲线进行智能化处理是非常需要但同时又是难于解决的一个问题。依据测井曲线特征可将测井处理解释知识分为基础信息、规则和策略,并结合解释参数、解释模型和解释图版等测井资料,构建了能够通过数据挖掘中的关联规则和项目集加权增量算法实现自动更新的测井处理解释知识库,基于该知识库构建了一个测井处理解释子系统,能够初步实现复杂储层测井处理解释从知识获取到参数和模型智能优选的全过程。通过在国家重大科技专项一体化平台CIFLog上的集成运行,证明了该处理流程及算法是科学有效的。

关键词: 知识库, 测井, 处理解释, 参数, 优选

Abstract:

During the evaluation of complex reservoir, optimal correcting chart and interpretation equation are selected automatically for target horizons with different property characteristics. It is not only quite necessary, but also a challenging issue for intelligent processing of logging curves. Based on logging curve characteristics, the knowledge of logging data processing and interpretation can be divided into basic information, rules and strategies. In combination with logging data such as interpretation parameters, models and charts, a knowledge base of logging processing and interpretation was established, which can update automatically using associated rules in data mining and item set weighting increment algorithm. On this basis, a relevant subsystem was created to preliminarily realize the whole process of logging data processing and interpretation for complex reservoirs, i.e., from knowledge acquisition to the intelligent optimization of parameters and models. Through integrated running on National Science and Technology Major Project Integration Software Platform (CIFLog), this processing and algorithm were proved to be scientific and effective.

Key words: knowledge base, logging, processing and interpretation, parameter, optimization

中图分类号: