Multivariate Control Charts Based on Principle Component Analysis and Partial Least Squares

The aim of this study is proposed new multivariate control charts by utilizing PCA technique feature for earlier detection of shifts in the process under Phase-II. So, to design new multivariate control charts based on PCA conventional multivariate MCUSUM and MEWMA control charts will be considered. Propose charts numerical results will be obtained through Monte Carlo simulation technique for comparison purpose against existing chart. The proposed charts will be evaluated using different famous performance measures such as average run length (ARL) for a shift, relative average run length (RARL), and some other related measures for overall assessment. Besides, to show the practical significance, the implementation of proposed will be demonstrated with real datasets from different industries such as health care, analytical and engineering.