Optimal maintenance policy and residual life estimation for a slowly degrading system subject to condition monitoring

• Published in: Reliability engineering & system safety Vol. 134; pp. 198 - 207
• Main Authors: Tang, Diyin, Makis, Viliam, Jafari, Leila, Yu, Jinsong
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.02.2015
• Subjects: Autoregressive model with time effect; Condition monitoring; Control-limit policy; Degradation; Inspection; Maintenance; Mathematical models; Optimization; Policies; Preventive maintenance; Residual life estimation; Semi-Markov decision process; Strategy; Condition monitoring; Autoregressive model with time effect; Control-limit policy; Residual life estimation; Semi-Markov decision process; Preventive maintenance
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2014.10.015

In this paper, we present an optimal preventive maintenance policy and develop a procedure for residual life estimation for a slowly degrading system subject to soft failure and condition monitoring at equidistant, discrete time epochs. An autoregressive model with time effect is considered to describe the system degradation, which utilizes both the system current age and the previous state observations. The class of control-limit maintenance policies with two different inspection strategies is considered, and the optimization problem is formulated and solved in a semi-Markov decision process framework. The objective is to minimize the long-run expected average cost. A formula for the mean residual life is derived for the proposed degradation model and a control limit policy, which enables the estimation of the remaining useful life and early maintenance planning based on the observed degradation process. An example is presented to demonstrate the effectiveness of the proposed method. •A new autoregressive model with time effect has been developed to model a slowly degrading process.•An optimal preventive maintenance policy has been developed.•A procedure for mean residual life estimation has been developed.•The whole procedure has been illustrated using a real data set.•Excellent results have been obtained.