流程性材料参数估计问题探讨  

A Research on the Parameter Estimator of Bulk Material

作　　者：孙小素[1] 戴金辉[1] 

机构地区：[1]山东工商学院统计学院,山东烟台264005

出　　处：《统计与信息论坛》2017年第7期49-53,共5页Journal of Statistics and Information

摘　　要：流程性材料最大的特点是其变异性小。对于这类总体,现有的标准差估计方法由于既包含组间差异,又包含组内差异,常常会夸大其估计误差。针对此,首先通过抽样设计,得到具有分层抽样特点的样本;然后借鉴单值—移动极差控制图中标准差的估计方法,构造了这类总体的标准差估计量。这样构造的标准差估计量,由于其估计误差中仅包含组内方差平均水平,从而更符合该类总体变异性小的特点。实际应用表明,该标准差估计量能显著降低估计误差。The distinguished characteristic of bulk materials is its small variability. The present estimation of the kind of population for its standard deviation will always be overstated for the deviation between and with-in stratum. We design a kind of stratified sampling so that the estimator for its standard deviation is only affected by the deviation with-in stratum. This idea comes from the inspiration in the individual-moving range control chart. The estimating method of the standard deviation for the designed by the author can cut down its variability significantly by the empirical application.

关 键 词：流程性材料 参数估计 分层抽样 单值—移动极差控制图 

分 类 号：O213.1[理学—概率论与数理统计] F224.0[理学—数学]