题目：Gaussian Process, Change-Point, and Remaining Useful Life Prediction

报告人：Rong Pan

报告时间：2024年4月28日（周日）9：30-10：30

地点：综合楼644

报告人介绍：

      Dr. Rong Pan is a Professor of Industrial Engineering and Data Science in the School of Computing and Augmented Intelligence (SCAI) at Arizona State University (ASU). He is the Program Chair of the Data Science, Analytics, and Engineering (DSAE) program at ASU. His research interests include failure time data analysis, design of experiments, multivariate statistical process control, time series analysis, and computational Bayesian methods. His research has been supported by NSF, Arizona Science Foundation, Air Force Research Lab, etc. He has published over 90 journal papers and 50+ refereed conference papers. Dr. Pan is a senior member of ASQ, IISE, and IEEE, and a lifetime member of SRE. He currently serves as the Chair of ASQ Reliability and Risk Division and the Editor-elect ofJournal of Quality Technology.

报告摘要：

      In this talk, a common production problem, in which frequent manufacturing equipment breakdowns and restarts cause low production efficiency, is presented. To predict a future breakdown time, we develop an early warning system that integrates Gaussian process (GP) modeling, change-point detection, online process monitoring, and remaining useful life prediction techniques. I will first introduce GP as a machine-learning tool for regression and classification tasks and then argue the values of derivative GP in change-point detection. To solve a practical engineering problem, we also employ the automatic relevance determination method for reducing feature dimensions. Finally, with a properly designed process monitoring tool, an early warning signal before equipment breakdown can be issued and the remaining equipment working time can be forecasted.