Abstract
There is much previous research on maintenance. Markov decision processes with unknown states are used in order to represent the maintenance problem. In the previous research a fixed profit is considered. In general the profit depends on economic states. In this research we propose a new maintenance method based on statistical decision theory. The proposed maintenance method maximizes the total reward of the maintenance problem considering profits and economic states with reference to a Bayes criterion. We show the effectiveness of the method by some numerical calculation examples. In the examples appropriate repairs are selected depending on economic states.
