State-age-dependent maintenance policies for deteriorating systems with Erlang sojourn time distributions

This paper investigates state-age-dependent maintenance policies for multistate deteriorating systems with Erlang sojourn time distributions. Since Erlang distributions are serial combinations of exponential phases, the deteriorating process can be modeled by a multi-phase Markovian model and hence easily analyzed. Based on the Markovian model, the optimal phase-dependent inspection and replacement policy can be obtained by using a policy improvement algorithm. However, since phases are fictitious and can not be identified by inspections, two procedures are developed to construct state-age-dependent policies based on the optimal phase-dependent policy. The properties of the constructed state-age-dependent policies are further investigated and the performance of the policy is evaluated through a numerical example.