Analysis of different maintenance policies on a multi-component system using dynamic bayesian networks

Recently, system components and interactions between them have become more complex and this situation has made it difficult to provide maintenance decisions. Herewith, determining effective decisions has played an important role. In multicomponent systems, many methodologies and strategies can be applied when a component or a system has already broken down or when it is desired to identify and avoid pro-actively defects that could lead to future failure. In dynamic systems, it is important for proactive maintenance to increase system reliability by performing early diagnosis-based maintenance activities without waiting for a problem. In this study, we focus on proactive maintenance of a complex multi-component dynamic system. Components are hidden although there exists partial observability to the decision maker. Components deteriorate in time. It is possible to replace or repair components with a given cost. We want to nd a policy that minimizes the total maintenance cost in a prede ned time horizon. We propose several maintenance policies and compare the performance of these by simulating them via Dynamic Bayesian Networks on an empirical model. Furthermore, a dynamic Bayesian network is constructed for the maintenance of an endo generator system to show how the proposed methods can be implemented in real life.