Multi-objective optimisation of multifaceted maintenance strategies for wind farms

Abstract

This paper proposes a way to simultaneously optimise three aspects of maintenance strategies for wind farms: the reliability thresholds that indicate when a component needs maintenance, the priority of maintenance jobs in case there are more jobs than available maintenance teams, and the use of opportunistic maintenance. We use a multi-objective evolutionary algorithm to simultaneously minimise the overall maintenance cost and the system idle time. An empirical study using a complex stochastic simulation demonstrates the benefit of such a multifaceted maintenance strategy over strategies that only consider some aspects.