A new multivariate statistical process monitoring method using principal component analysis

Principal component analysis (PCA) has been used successfully as a multivar iate statistical process control (MSPC) tool for detecting faults in proces ses with highly correlated variables. In the present work, a novel statisti cal process monitoring method is proposed for further improvement of monito ring performance. It is termed 'moving principal component analysis' (MPCA) because PCA is applied on-line by moving the time-window. In MPCA, changes in the direction of each principal component or changes in the subspace sp anned by several principal components are monitored. In other words, change s in the correlation structure of process variables, instead of changes in the scores of predefined principal components, are monitored by using MPCA. The monitoring performance of the proposed method and that of the conventi onal MSPC method are compared with application to simulated data obtained f rom a simple 2 x 2 process and the Tennessee Eastman process. The results c learly show that the monitoring performance of MPCA is considerably better than that of the conventional MSPC method and that dynamic monitoring is su perior to static monitoring, (C) 2001 Elsevier Science Ltd. All rights rese rved.