Maintenance optimization considering the mutual dependence of the environment and system with decreasing effects of imperfect maintenance

A system state and its working environment can be mutually dependent, such as when the surrounding area is relatively confined, and there is a strong interaction between them. Maintenance strategies without considering this relationship can be ineffective. In this study, a novel maintenance strategy is developed for multi-state systems operating exposed to changing environment states, and the system and environment can influence each other. Also, the operation and maintenance costs, and the maintenance effects are influenced by the current system and environment. In this study, three actions are selectable at decision-making times, including do nothing, imperfect maintenance and replacement. Imperfect maintenance can improve the system state to some extent, but not fully. Then subsequently, the environment state also improves, which occurs with a time delay after the system improvement. In addition, the effect of imperfect maintenance decreases as the number of completed imperfective maintenance activities increases. Whenever the system is at its worst state, replacement is required. A Markov decision process is formulated to model the problem. By using a policy iteration algorithm, the maintenance strategy can then be optimized. Numerical examples are given to demonstrate the proposed method and its effectiveness, and sensitivity analyses of parameters are also performed.

Automated summary

In this study, a novel maintenance strategy is developed for multi-state systems operating exposed to changing environment states, and the system and environment can influence each other. A Markov decision process is formulated to model the problem, and a policy iteration algorithm is used to optimize the maintenance strategy. Numerical examples are provided to demonstrate the effectiveness of the proposed method, along with sensitivity analyses of parameters.