最大似然法储层预测技术研究及软件研制  

RESEARCH OF THE MAXIMUM LIKELIHOOD METHOD FOR RESERVOIR PREDICTION AND SOFTWARE DEVELOPMENT

作　　者：董建萍[1] 董国强[1] 勾永峰[1] 

机构地区：[1]大庆钻探工程公司地球物理勘探一公司,黑龙江大庆163357

出　　处：《石油工业计算机应用》2009年第1期14-15,共2页Computer Applications Of Petroleum

摘　　要：本文阐述了最大似然法储层预测软件的原理及应用情况。通常储层预测采用的是多元线性回归和人工神经网络等方法,很少用到最大似然法进行储层预测的,而最大似然法分类能力比较强,可以由单井出发进行多井外推预测,从而较精确地预测了储层和油气情况。通过在实际生产工区的试验应用,该方法可以将整个工区清楚地分类,对提高储层油气预测精度,具有很大的帮助作用。The paper formulates the principle and application of the software for reservoir prediction with maximum likelihood method.Multielement linear regression and artificial neural network are usually used in reservoir prediction,but maximum likelihood method is seldom used.It has an advantage in classification which can be extrapolated to multi-wells prediction from a single well.Hydrocarbon in reservoirs can be predicted more accurately with this method.Through experiment application in the work area,it is proved that the whole work area can be classified clearly with this method.It is helpful for improving the prediction precision of reservoir.

关 键 词：最大似然法 地震属性 储层预测 

分 类 号：P618.13[天文地球—矿床学] TN911.23[天文地球—地质学]