Dynamic preventive maintenance policy based on health index

The purpose of this article is to propose a dynamic preventive maintenance policy for the repairable deteriorating system. Assume that the system after repair can not be "as good as new", and general repair which brings nonstationary recovery is adopted. We propose the health index (HI) which can be inspected by equipment to describe the status of the system, and suppose that general repair has the corresponding recovery ability for different health index. Then, a discrete Markov decision process (MDP) for the maintenance policy optimization under minimizing the long-term average cost rate is built. Finally, a real case study from Chen and Feldman (1997) is presented to verify that the abovementioned methodology is efficient and practical.