Performance-oriented risk evaluation and maintenance for multi-asset systems: A Bayesian perspective

Abstract

In this article, we present a risk evaluation and maintenance strategy optimization approach for systems with parallel identical assets subject to continuous deterioration. System performance is defined by the number of functional assets, and the penalty cost is measured by the loss of performance. To overcome the practical challenges of information sparsity, we employ a Bayesian framework to dynamically update unknown parameters in a Wiener degradation model. Order statistics are utilized to describe the failure times of assets and the stepwise incurred performance penalty cost. Furthermore, based on the Bayesian parameter inferences, we propose a short-term value-based replacement policy to minimize the expected cost rate in the current planning horizon. The proposed strategy simultaneously considers the variability of parameter estimators and the inherent uncertainty of the stochastic degradation processes. A simulation study and a realistic example from the petrochemical industry are presented to demonstrate the proposed framework.

KEYWORDS:

• Bayesian inference
• order statistics
• parameter uncertainty
• performance evaluation
• preventive maintenance

Acknowledgments

We are grateful to the editors and anonymous referees for their helpful comments and suggestions on an earlier version of the paper.

Funding

This work was supported in part by the National Natural Science Foundation of China (grant numbers 72002149, 72001124, 71971181, 72032005) and the Research Grants Council of Hong Kong (GRF 11203519).

Additional information

Notes on contributors

Xiujie Zhao

Xiujie Zhao is an associate professor with the College of Management and Economics, Tianjin University, Tianjin, China. He received a BE degree from Tsinghua University, an MS degree from the Pennsylvania State University and a PhD degree from City University of Hong Kong, all in industrial engineering. His research interests include accelerated reliability testing, degradation modeling, maintenance optimization and design of experiments. His papers have appeared in IISE Transactions, European Journal of Operational Research, Journal of Quality Technology, IEEE Transactions on Reliability, Reliability Engineering & System Safety, among others.

Zhenglin Liang

Zhenglin Liang received his PhD at the University of Cambridge, St John’s College. He is currently an assistant professor in system engineering, Department of Industrial Engineering, at Tsinghua University. His research interests include: system reliability, predictive maintenance, machine learning and modelling of complex networks.

Ajith K. Parlikad

Ajith K. Parlikad is reader in asset management at the Department of Engineering in the University of Cambridge. Ajith’s research focuses on exploiting data and digital technologies to improve the maintenance and operation of industrial and infrastructure systems. He is the Scientific Secretary of the IFAC Working Group on Advanced Maintenance Services and Technology and sits on the Steering Board of the UK Digital Twin Hub.

Min Xie

Min Xie received his PhD from Linkoping University, Sweden in 1987. He did his undergraduate study and received an MSc at Royal Institute of Technology in Sweden in 1984. Dr. Xie joined the National University of Singapore in 1991 as one of the first recipients of the prestigious Lee Kuan Yew Research Fellowship, and he is currently a Chair Professor at the City University of Hong Kong. He has authored or co-authored numerous refereed journal papers and several books. He is a Department Editor of IISE Transactions and Editor of Reliability Engineering & System Safety and serves in a number of other international journals. He has organized many international conferences, and also 50 PhD students have graduated under his supervision. He was elected fellow of IEEE in 2005.