基于支持向量机的属性优选和储层预测

英文篇名：Attributes selection and reservoir prediction based on support vector machine
中文刊名：石油地球物理勘探
英文刊名：Oil Geophysical Prospecting
作者：张长开 ; 姜秀娣 ; 朱振宇 ; 印海燕 ; 陆文凯
英文作者：Zhang Changkai1 ; Jiang Xiudi2 ; Zhu Zhenyu2 ; Yin Haiyan2 and Lu Wenkai1.1.Department of Automation ; State Key Laboratory of Intelligent Technology and Systems ; Tsinghua National Laboratory for Information Science and Technology ; Tsinghua University ; Beijing 100084 ; China2.CNOOC Research Institute ; Beijing 100027 ; China
中文关键词：支持向量机 ; 地震属性 ; 属性优选 ; 支持向量回归机 ; 储层预测
英文关键词：support vector machine ; seismic attributes ; attribute selection ; support vector regression(SVR) ; reservoir prediction
出版日期：2012-04-15
机构：清华大学自动化系智能技术与系统国家重点实验室清华大学信息科学与技术国家实验室(筹);中海油研究总院;
年：2012
期：02
出版单位：石油地球物理勘探

摘要

本文采用基于支持向量机(SVM)的特征选择方法进行地震属性优选,根据油井的产油气情况将油井分为高产井和低产井,利用SVM对这些样本进行训练,然后根据每个属性对应的权值进行筛选,便可以选取对油气敏感的属性,进而更好地预测储层。具体过程为:①提取一定量的地震属性;②根据已知井的信息,获得训练样本,训练线性SVM;③计算各个特征的权值;④选取较大权值绝对值对应的多个属性;⑤将支持向量回归机(SVR)应用于优选出的属性,获得储层预测的结果。实际资料应用结果表明,文中方法不仅能筛选出有效的地震属性,还能够有效地预测储层。
This paper applies feature selection algorithm based on SVM to select seismic attributes.According to the oil and gas yielding of oil wells,samples of seismic attributes are divided into two kinds:high-yielding well and low-yielding well.After these samples are trained by SVM,the attributes sensitive to oil and gas will be selected by screening the weight corresponding to each attribute,and then be taken advantage to predict reservoirs.The detailed process can be described as:①Extract certain seismic attributes;②Obtain samples according to the given information of some oil wells and train them by using SVM;③Calculate the weight of every attribute;④Choose out the attributes whose weight absolute value are respectively large;⑤Apply the support vector regression(SVR) to the chosen out attributes and predict reservoir.Our application of this algorithm on real seismic data shows that the algorithm is able to choose out valid seismic attributes and effectively predict reservoirs at the same time.

引文

[1]王咸彬,顾石庆.地震属性的应用与认识.石油物探,2004,43(增刊):25～27
    [2]姜岩,李文艳,吴明华.一种模糊神经网络技术及其在储层预测中的应用.石油物探,2004,43(4):377～379Jiang Yan,Li Wenyan and Wu Minghua.Fuzzy neu-ral network and its application in reservoir prediction.GPP,2004,43(4):377～379
    [3]杨培杰,印兴耀,张广智.模糊C均值地震属性聚类分析.石油地球物理勘探,2007,42(3):322～324,327Yang Peijie,Yin Xingyao and Zhang Guangzhi.Clus-ter analysis of seismic attributes by fuzzy C-mea algo-rithm.OGP,2007,42(3):322～324,347
    [4]赵虎,尹成,朱仕军.多属性融合技术研究.勘探地球物理进展,2009,32(2):119～122Zhao Hu,Yin Cheng,Zhu Shijun.Technique of seis-mic multi-attributes fusion.PEG,2009,32(2):119～121
    [5]Vapnik V著,张学工译.统计学习理论的本质.北京:清华大学出版社,2000
    [6]张学工.关于统计学习理论与支持向量机.自动化学报,2000,26(1):32～41Zhang Xuegong.Introduction to statistical learningtheory and support vector machines.Acta Automati-ca Sinica,2000,26(1):32～41
    [7]金龙,陈小宏,王守东.基于支持向量机与信息融合的地震油气预测方法.石油地球物理勘探,2006,41(1):76～80Jin Long,Chen Xiaohong and Wang Shoudong.Oil/gas-predicted method by seismology based on sup-porting vector computer and information merging.OGP,2006,41(1):76～80
    [8]唐耀华,张向君,高静怀.基于地震属性优选与支持向量机的油气预测方法.石油地球物理勘探,2009,44(1):75～80Tang Yaohua,Zhang Xiangjun and Gao Jinghua.Method of oil/gas prediction based on optimization ofseismic attributes and support vector machine.OGP,2009,44(1):75～80
    [9]姚凯丰,陆文凯,丁文龙等.一种基于SVM特征选择的油气预测方法.天然气工业,2004,24(7):36～38Yao Kaifeng,Lu Wenkai,Ding Wenlong et al.Hy-drocarbon prediction method based on SVM featureselection.Natur Gas Ind,2004,24(7):36～38
    [10]印兴耀,周静毅.地震属性优化方法综述.石油地球物理勘探,2005,40(4):284～289Yin Xingyao and Zhou Jingyi.Summary of optimummethods of seismic attributes.OGP,2005,40(4):284～289
    [11]季敏,王尚旭,陈双全.地震属性优选在油田开发中的应用.石油地球物理勘探,2006,41(2):183～187Ji Min,Wang Shangxu and Chen Shuangquan.Appli-cation of optimum seismic attributes in developmentof oilfiled.OGP,2006,41(2):183～187