基于最小二乘支持向量机的重复压裂选井选层方法研究  

Research on method of Repeated Fracturing to Choose Well and Layer Based on Least Squares Support Vector Machines

作　　者：宋时权[1] 

机构地区：[1]胜利油田采油工艺研究院,山东东营257000

出　　处：《油气井测试》2013年第4期13-14,20,共3页Well Testing

基　　金：中国石油天然气总公司重大专项研究项目(CNPC 2007-20071209)

摘　　要：常用的重复压裂选井选层的经验方法是基于矿场统计资料的研究方法,这种方法具有一定的主观性、盲目性和风险性。引入基于统计学习理论的最小二乘支持向量机算法,综合考虑地层地质特征、油气藏特性、物性参数、测试和生产数据等多方面的因素,形成重复压裂选井选层方法。计算证明,该方法对重复压裂选井选层有指导作用。The common experience of the refracturing Wells of selected layer method is based on field statistical data research method,this method has a certain subjectivity,blindness and risk.Introduced based on the statistical learning theory of least squares support vector machine(SVM)algorithm,considering reservoir geological characteristics,reservoir characteristics,physical parameters,test and production data,and so on various factors,the formation of refracturing wells picked layer method.Calculations show that the method of repeated fracturing to choose well layer has a guiding role.

关 键 词：重复压裂 选井选层 最小二乘支持向量机 

分 类 号：TE353[石油与天然气工程—油气田开发工程]