欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
多变异分析将过程质量特征值的变异分为时间变异、产品间变异和产品内变异。产品间变异和时间变异的概率分布均为卡方随机变量的线性组合，而这些线性组合的分布表达式尚未求出，所以这两种变异的方差分量不能给出确切的置信区间。在介绍了求方差分量近似置信区间方法的基础上，将其与多变异分析的统计模型相结合，给出了产品间变异和时间变异分量的近似置信区间的求法。
英文摘要：
In multi-vari analysis （ MVA）, the variation sources of product/process quality characteristics are divided into time to time variation, piece to piece variation and within piece variation. But for the time to time and the piece to piece vafiance components, an expression for the distribution of their linear combination of chi-square random variables cannot be obtained. Thus, an exact confidence interval for them cannot be constructed. In this paper, approximate procedures are given. Based on the statistical model of MVA and the approximate procedures, the approximate confidence intervals of these two variance components are discussed.
同期刊论文项目
同项目期刊论文
位置：