Applied Stochastic Models in Business and Industry, vol.36, no.3, pp.381-396, 2020 (SCI-Expanded, Scopus)
In this study, a mechanical system with linear deterioration and preventive maintenance is considered. The state of the system over time is represented by a semicontinuous stochastic process with dependent components. The system cycles through on and off periods during its lifetime. The state of the system deteriorates linearly as a function of the usage time during on periods. When the system is offline, preventive maintenance is conducted, which improves the system state by a random amount. The system's on and off times and random improvement amounts are assumed to have general distributions. For such a system, our objective is to determine the expected value and variance for the number of preventive maintenance activities needed during the system lifetime and to propose a novel replacement policy for the system based on delay-time modeling. Finally, the effectiveness of the obtained asymptotic results and the proposed replacement policy are tested through simulation.