论文部分内容阅读
王瑞平,白康生,赵文杰,等.基于数理统计理论的两种分层方法的比较.测井技术,1998,22(5):332~336随着地下资源的枯竭和测井技术的发展,人们期待着更精确更可靠地寻找到储集层的方法。介绍并利用费歇尔(Fisher)判据分类方法识别划分目的层,并把这种方法与马哈诺比斯(Mahalanobis)距离分层方法加以比较。两种方法的联合应用使分层具有“可比性”。两种方法依据多元统计理论,对“标准”传统分层资料进行随机抽样,由此可进行误差分析,判断资料的可靠性和方法的实用性。对胜利油田某口井作了实际处理,效果较好。
Wang Ruiping, Bai Kangsheng, Zhao Wenjie, and so on. Comparison of Two Kinds of Layered Methods Based on Mathematical Statistics. Well Logging Technology, 1998,22 (5): 332 ~ 336 With the depletion of underground resources and the development of well logging technology, people are looking forward to finding reservoirs more accurately and reliably. This paper introduces and uses the Fisher criterion classification method to identify the target stratification and compares this method with the Mahalanobis distance stratification method. The combined use of the two approaches makes stratification “comparable.” According to the multivariate statistical theory, the two methods randomly sample the “standard” traditional stratified data, which can be used to analyze the error and judge the reliability of the data and the practicability of the method. A well in Shengli Oilfield made a practical treatment, the effect is better.