High Dimensional Process Monitoring Using Principle Component Analysis and T2 chart

Jalilibal         Z. | Mousavi         S.M. | Amirhossein Amiri

---

URL :   http://research.shahed.ac.ir/WSR/WebPages/Report/PaperView.aspx?PaperID=127111
Date :  2019/09/04
Publish in :    5th International Conference on Industrial and Systems Engineering


Keywords :Principal component analysis, High dimensional process monitoring, dimension reduction, Hotelling ????.

Abstract :

---

Statistical process monitoring is an essential need for industrial processes. Many of these processes apply principal component analysis to perform statistical process monitoring as its simplicity in computations. The PCA is used in this study to reduce dimension for monitoring high dimensional process which has complex computations. Fault detection charts that are commonly employed with the PCA method are the Hotelling 𝑻𝟐 statistic which are used for monitoring the process which is reduced by PCA. This study has two steps; first, the high dimensional process is reduced by applying PCA, and then, the reduced process is monitored.