A case study of condition-based maintenance modeling based upon the vibration analysis data of wind machines

This paper presents a maintenance optimization model based upon the vibration data of the wind machines in a steel mill. The wind machines in a steel mill require regular rebalancing and eventual overhaul to avoid catastrophic failure. The decision model for minimizing the expected maintenance cost per unit time of the wind machines is derived conditional on the number of rebalances within an overhaul cycle. We use a Wiener process to model the degradation process of the wind machines and the first time to reach the rebalancing threshold is modeled by the first passage time. Finally, a simulation algorithm is presented to evaluate the expected maintenance cost per unit time.