Optimal non-periodic replacement and reactivation in standby systems with protection and maintenance options

Abstract

Traditional standby system models have mostly assumed that each element can be online and used once during the mission time. In practice especially when a long mission task is performed, it is beneficial to reactivate some used elements to enhance the mission success probability. This paper proposes a new standby system model, where non-periodic preventive replacements (PR) are conducted to mitigate operating elements’ accumulated deterioration and an element may be activated several times to be online executing the mission task. Moreover, each used element undergoes idle time dependent protection or maintenance actions before its next reactivation. The mission is successful if during the specified mission time no activated element fails while online and operating or before its activation. The mission success probability is evaluated using a probabilistic approach proposed in this paper. The optimal PR scheduling problem is further formulated and solved, which determines the element activation/reactivation schedule maximizing the mission success probability. Examples of homogeneous and heterogeneous standby systems are provided to examine influences of several model parameters and their interactions on the mission success probability as well as on optimal PR schedules. The application of the proposed model to the element sensitivity analysis is also illustrated through example analysis.