## Journal of Central South University

 第46卷    第6期    总第250期    2015年6月

[PDF全文下载]    [Flash在线阅读]

(1. 中国石油大学 地球科学与技术学院，山东 青岛，266580；
2. 中国石油大学CNPC测井重点实验室，山东 青岛，266580；
3. 中国石油大庆油田井下作业分公司，黑龙江 大庆，163000；
4. 陕西延长石油(集团)有限责任公司研究院，陕西 西安，710075；
5. 中国石油大庆油田海拉尔石油勘探开发指挥部，黑龙江 大庆，163000；
6. 中国石油集团测井有限公司长庆事业部，陕西 西安，710201
)

Pore structure typing of heterogeneous clastic reservoir using information entropy-fuzzy spectral clustering algorithm

1. School of Geosciences in China University of Petroleum, Qingdao 266580, China;
2. CNPC Key Well Logging Laboratory in China University of Petroleum, Qingdao 266580, China;
3. Down Hole Service Sub-Company of Daqing Oilfield Company, Daqing 163000, China;
4. Research Institute of Shanxi Yanchang Petroleum (Group) Co. Ltd., Xi’an 710075, China;
5. Hailar Headquaters of Petroleum Exploration and Development of Daqing Oil field Company, Daqing 163000, China;
6. Changqing Division, China Petroleum Logging Co. Ltd., Xi’an 710201, China

Abstract:A method of automatic typing was proposed by using information entropy-fuzzy spectral clustering algorithm. Complete convergence was obtained by using spectral clustering algorithm to solve the convex distribution clustering problem, thus the ‘dimension disaster’ was avoided effectively. In the light of information entropy theory, the scale parameters of spectral clustering algorithm were optimized and then the types of pore structures were presented. On this basis, in combination with each sample of the pore structure types of membership from fuzzy mathematics algorithm, various types of pore structure of different samples can be obtained according to the membership degree of optimal rule (membership degree of samples for one pore structure is greater than 70%). The results show that the results of pore structure typing obtained from the algorithm are in good agreement with well test and production test results, and its engineering application effect is very obvious.

Key words: heterogeneous clastic reservoir; pore structure typing; fuzzy spectral clustering algorithm; information entropy; scale parameter optimizing

 中南大学学报（自然科学版） ISSN 1672-7207 CN 43-1426/NZDXZAC 中南大学学报（英文版） ISSN 2095-2899 CN 43-1516/TBJCSTFT