地震属性奇异值分解法降维的应用  

THE SEISMIC ATTRIBUTES APPLICATION OF SINGULAR VALUE DECOMPOSITION METHOD FOR DIMENSIONALITY REDUCTION

作　　者：刘邦兴[1] 曾良雄[1] 岳红星[1] 

机构地区：[1]中国石油新疆油田公司准东采油厂,新疆阜康831511

出　　处：《新疆石油天然气》2015年第2期25-27,43,共4页Xinjiang Oil & Gas

摘　　要：在地震属性领域中,从地震资料中提取的地震属性可达上百种,大多情况是凭解释人员的经验来完成地震属性挑选的,不但增加了工作量,人为因素也难以避免,为解决这一困难,需要把数据挖掘技术引入地震属性领域。数据降维就是数据挖掘的一种,本文将奇异值分解降维方法应用到地震属性降维中,通过与钻井、试油的对比结果来看,在储层含油气预测方面数据降维方法是一种有效的分析方法。In the field of seismic attribute,seismic attributes can be extracted from seismic data up to hundreds,mostly explained by people's experience to complete the seismic attribute selection,which not only increases the workload,but the human factor is difficult to avoid,so the data mining technology is introduced into the field of seismic attributes,one of which is data mining dimension data reduction.In this paper,the singular value decomposition dimension reduction method is applied to the seismic attribute reduction,by comparing the results with drilling,test oil,this method is an effective method in oil / gas reservoir.

关 键 词：地震属性 数据降维 奇异值分解 

分 类 号：P631[天文地球—地质矿产勘探]