Abstract: Selecting the multi-component system of wind turbine as object of study, aiming at its characteristics of difficulty in maintenance and heavy downtime loss, the dynamic opportunistic maintenance and its optimization for multi-component system is carried out. The age reduction factor is introduced to describe the effect of imperfect maintenance, the average maintenance cost model of single component is proposed on the basis of the definition of failure rate function. Considering the economic dependence among the components, and by taking into account the cost of preventive maintenance, minor maintenance and downtime, the maintenance decision model in a finite time horizon is presented, which aims to maximize the total opportunistic maintenance cost saving by arranging the maintenance activities. In view of the harsh and changing operating environment of wind turbines, dynamic maintenance scheduling is realized based on rolling horizon approach, the stochastic downtime event is used to arrange the maintenance activities and the maintenance plan is updated timely. Case study of a wind turbine including four key components of gearbox, rotor, generator and bearing is provided. The result shows that the presented model can update the maintenance plan in time so as to dynamically optimize the maintenance activities and reduce the maintenance cost.

Key words: wind turbine;preventive maintenance;opportunistic maintenance;rolling horizon approach;dynamic optimization