储层综合识别智能模型的建立及在准噶尔盆地西北缘的应用  

Establishment of intelligent model for comprehensive identification of reservoir and application of it in the northwestern margin of Jungger Basin

作　　者：郭大立[1] 许江文 贾海正 柯西军[1] 李雪彬[3] 

机构地区：[1]西南石油大学理学院,四川成都610500 [2]新疆油田公司工程技术研究院,新疆克拉玛依834000 [3]新疆油田公司勘探公司,新疆克拉玛依834000

出　　处：《西安石油大学学报（自然科学版）》2015年第6期41-44,8,共4页Journal of Xi’an Shiyou University（Natural Science Edition）

基　　金：国家科技重大专项示范工程(编号:2011ZX05062-008)

摘　　要：为准确识别储层类型,针对准噶尔盆地西北缘三叠系储层的特点,提出了一种结合人工神经网络专家和模糊聚类分析的综合识别储层方法。该方法先用10个专家(用BP神经网络获取的知识)进行识别,如果有一半或者一半以上的专家识别结果一致,则得到识别结果,否则采用模糊聚类分析模型进行识别。对西北缘三叠系的17个层进行识别,符合率为76.47%。In order to accurately identify the type of reservoir,a comprehensive reservoir type identification method is presented according to the characteristics of Triassic reservoir in the northwestern margin of Jungger Basin,in which artificial neural network experts are combined with fuzzy cluster analysis. Ten experts( knowledge acquired by BP neural network) identify firstly,and if half or more of10 experts get the same result,then the comprehensive identification result is obtained; if not,the identification is finished using fuzzy cluster analysis model. The comprehensive reservoir type identification method is applied to 17 layers in the Triassic reservoir in the northwestern margin of Jungger Basin,the coincidence ratio of the identification results reaches to 76. 47%.

关 键 词：储层综合识别 模糊聚类分析 神经网络 三叠系储层 准噶尔盆地西北缘 

分 类 号：TE357.1[石油与天然气工程—油气田开发工程]