PCA-based methods for monitoring high-dimensional, time-dependent processes

This dissertation will focus on investigating and improving techniques used in the statistical process control field, and particular with regards to control charts, as they pertain to the control of systems which display autocorrelation, non-stationarity, and exhibit outlier and missing measurements. Control chart approaches which will initially be considered are Dynamic PCA/PLS, MovingWindow PCA/PLS, and Recursive PCA/PLS. Each of these approaches attempts to extend the static control chart model so that it may be applied in a setting with time series dynamics. As of yet, no comprehensive overview of the methods and their extensions has been performed to evaluate their relative performance. Furthermore, in all cases, additional innovations are needed before these methods display the straightforward applicability of static control charts. Therefore, (1) examining the characteristics of the various approaches discussed in the literature, and (2) providing innovative extensions which yield stable and easy to use methods will form two primary research objectives.status: publishe