Selective Maintenance Optimization for a Multi-State System With Degradation interaction

This paper studies the selective maintenance problem for a multi-state system (MSS) performing consecutive production missions with scheduled intermission breaks. To improve the reliability of the system successfully performing the next mission, all maintenance actions need to be carried out during maintenance breaks. However, it may not be feasible to repair all components due to the limited maintenance resources (such as time, costs, and manpower). Hence, a selective maintenance model was established to identify a subset of maintenance actions to perform on the repairable components. We extend the original model in several ways. First, we consider the role of degradation interaction in determining the state transition probability of each component. Back-propagation (BP) neural network is employed to predict the transition matrix since it is not practicable to analyze the degradation processes of all components using the traditional probability model. Second, a selective maintenance optimization model for an MSS is established based on the prediction results of the BP neural network and solved by a genetic algorithm (GA). Finally, an example is illustrated to verify the effectiveness and superiority of the proposed method.