Remaining useful life prediction based on the mixed effects model with mixture prior distribution

ABSTRACT

Modern engineering systems are gradually becoming more reliable and premature failure has become quite rare. As a result, degradation signal data used for prognosis are often imbalanced as most units are reliable and only few tend to fail at early stages of their life cycle. Such imbalanced data may hinder accurate Remaining Useful Life (RUL) prediction especially in terms of detecting premature failures as early as possible. This aspect is detrimental for developing cost-effective condition-based maintenance strategies. In this article, we propose a degradation signal–based RUL prediction method to address the imbalance issue in the data. The proposed method introduces a mixture prior distribution to capture the characteristics of different groups within the same population and provides an efficient and effective online prediction method for the in-service unit under monitoring. The advantageous features of the proposed method are demonstrated through a numerical study as well as a case study with real-world data in the application to the RUL prediction of automotive lead–acid batteries.

KEYWORDS:

• Degradation signals
• imbalanced data
• mixture prior
• remaining useful life

Funding

This research was supported by the National Science Foundation under grant number 1335129.

Additional information

Notes on contributors

Raed Kontar

Raed Kontar is a Ph.D. candidate in Industrial and Systems Engineering and an M.S. candidate in Statistics, both at the University of Wisconsin–Madison. He received a B.S. in Civil and Environmental Engineering (2014) from the American University of Beirut, Lebanon.

Junbo Son

Junbo Son is an Assistant Professor of Operations Management at Alfred Lerner College of Business & Economics, University of Delaware. He received a B.S. in Industrial Engineering (2010) from Korea University, South Korea; an M.S. in Statistics (2015); and a Ph.D. in Industrial and Systems Engineering (2016), both from the University of Wisconsin–Madison. His research interests are data analytics and data-driven decision making in complex manufacturing, service, and healthcare systems. He is a member of IISE and INFORMS.

Shiyu Zhou

Shiyu Zhou is a Professor in the Department of Industrial and Systems Engineering at the University of Wisconsin–Madison. He received his B.S. and M.S. in Mechanical Engineering from the University of Science and Technology of China in 1993 and 1996, respectively, and his master's in Industrial Engineering and Ph.D. in Mechanical Engineering from the University of Michigan in 2000. His research interests include in-process quality and productivity improvement methodologies by integrating statistics, system and control theory, and engineering knowledge. He is a recipient of a CAREER Award from the National Science Foundation and the Best Application Paper Award from IIE Transactions. He is a member of IIE, INFORMS, ASME, and SME.

Chaitanya Sankavaram

Chaitanya Sankavaram is a Senior Researcher with the Vehicle Systems Research Lab at General Motors Global R&D Center in Warren, Michigan. She joined General Motors in 2013 and since then she has been involved in the research and development of prognostics and health management algorithms for automotive systems. She received her B.Tech. degree in Electrical and Electronics Engineering from Sri Venkateswara University, Tirupathi, India, and M.S. and Ph.D. degrees in Electrical and Computer Engineering from the University of Connecticut, Storrs, in 2013 and 2015, respectively. Her research interests include fault diagnosis and prognosis, machine learning, data mining, pattern recognition, reliability analysis, and optimization theory. She was a co-recipient of the Walter E. Peterson Best New Technology paper award from IEEE Autotestcon in 2011. She also received a Connecticut Women of Innovation finalist award from the Connecticut Technology Council in 2012. She has published over 20 peer-reviewed technical publications and is a member of IEEE, Sigma Xi, and the Society of Women Engineers.

Yilu Zhang

Yilu Zhang is the manager of the Vehicle Health Management Group at General Motors Global R&D Center, Warren, Michigan. He received his Ph.D. degree in Computer Science from Michigan State University, East Lansing, Michigan. His research interests include statistical pattern recognition, machine learning, signal processing, and their applications, including integrated vehicle health management and human machine interactions. He served as an Associate Editor of the International Journal of Humanoid Robotics from 2003 to 2007, the Publication Chair for the IEEE Eighth International Conference on Development and Learning 2009, and the Chair of the Battery Management System Workshop in conjunction with PHM Society Annual Conference 2011. He has published 50+ refereed technical papers. He has 29 U.S. patents and 20+ pending patent applications, of which most are in the area of vehicle diagnosis and prognosis. He is a three-time recipient (2008, 2010, 2015) of the “Boss” Kettering Award, the highest technology award in General Motors, for his contribution to vehicle diagnostics and prognostics technologies. He is a senior member of IEEE.

Xinyu Du

Xinyu Du received B.Sc. and M.Sc. degrees in Automation from Tsinghua University, Beijing, China, in 2001 and 2004, respectively, and a Ph.D. degree in Electrical Engineering from Wayne State University, Michigan, in 2012. He has been working at the General Motors Global R&D Center, Warren, Michigan, since 2010 and currently holds a senior researcher position in the Vehicle Systems Research Lab. He has published 27 peer-reviewed papers and holds four patents. He has 11 U.S. or international patents pending and seven GM internal inventions. His research interests include fuzzy hybrid systems, vehicle health management, vehicle electronics, and data mining. He has served as an Associate Editor for the Journal of Intelligent and Fuzzy Systems since 2012 and has served as a lead guest editor for Advances in Fuzzy Systems. He was a member of the IEEE CIS GOLD sub-committee in 2011 and 2012 and a member of IEEE Fuzzy System Competitions Technical Committee in 2009 and 2011. He received the best student paper finalist award from the 24th NAFIPS annual conference in 2005, the Ralph H. Kummler Award for distinguished achievement in graduate student research from Wayne State University in 2010, and the Boss Kettering Award by General Motors in 2015 for his contributions to integrated starting system prognosis.