(101d) Quality-Relevant Process Monitoring | AIChE

(101d) Quality-Relevant Process Monitoring


Qin, S. J. - Presenter, University of Southern California

Data-driven process monitoring based on principal component analysis (PCA) and partial least squares (PLS) has received wide acceptance in industry during the past two decades. Extensions are developed in various directions to deal with specific characteristics of process and quality data. In the paper we present recent results in new understanding of the use of PLS-like methods for quality-relevant process monitoring. The PLS-based monitoring is an area that is less studied and understood compare to PCA-based monitoring methods. We discuss new ways of partitioning the process data space and defining quality relevant and irrelevant monitoring indices. Fault diagnosis methods are discussed in this context as well.