Abstract
Advanced sensing technology enables the collection of multi-channel profiles in real-time, facilitating in-process monitoring and anomaly detection. This paper presents an in-profile monitoring (INPOM) control chart for cluster-correlated data, addressing issues such as detection delay and sample misalignment. A regularized state space model (RSSM) is proposed to describe the data, with model parameters learned via an EM algorithm. The effectiveness of the RSSM-INPOM framework is demonstrated through numerical studies and real case applications.
Juan Du
Assistant Professor
My research interests include knowledge-infused data science for quality improvement, industrial data analytics and machine learning, and system informatics and control for manufacturing applications.