论文部分内容阅读
作为多元统计过程控制方法中的常用统计量，平方预测误差（ Ｓ Ｐ Ｅ）的变化规律有待深入研究。介绍了主元分析建模方法，推导了 Ｓ Ｐ Ｅ均值公式，分析了 Ｓ Ｐ Ｅ均值和过程变量均值向量、协方差矩阵之间的解析关系，用来自 ３ 阶液位系统的仿真数据验证了分析的结果。给出了 Ｓ Ｐ Ｅ随过程变量均值向量、协方差矩阵变化而变化的若干规律，说明了这些规律在生产过程监控应用中的意义。
As a commonly used statistical measure in multivariate statistical process control methods, the variation rule of squared prediction error (S P E) needs further study. The principal component analysis modeling method is introduced and the mean value formula of S P E is deduced. The analytic relationship between mean value of S P E and mean value vector of process variables and covariance matrix is analyzed. The simulation data from the third order liquid level system is validated Results of the analysis. Some rules that S P E vary with the vector and covariance matrix of process variables are given. The significance of these rules in the monitoring of production process is illustrated.