摘要
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。
Squared prediction error (SPE) statistic is frequently used in multivariate statistical process control and its law of variation needs further investigating. The modelling method of principal component analysis (PCA) is introduced. The mean equation of SPE is developed. The analytical relationships of SPE mean and process variables mean and covariance matrix are analyzed. Simulation data from a 3rd order level system are used to validate the results obtained. Some laws of variation of the SPE mean under the variation of mean and covariance matrix of process variables are given. The significance of the laws in the application of process monitoring is explained.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
1999年第7期41-45,共5页
Journal of Tsinghua University(Science and Technology)
基金
国家"八六三"高技术研究课题
关键词
多元统计
过程控制
平方预测误差
工业控制
multivariate statistical process control
squared prediction error (SPE)
principal component analysis
process monitoring
