Advanced signal processing techniques for fault detection and diagnosis in a wind turbine induction generator drive train: A comparative study

Abstract

This paper deals with the diagnosis of Wind Turbines based on generator current analysis. It provides a comparative study between traditional signal processing methods, such as periodograms, with more sophisticated approaches. Performances of these techniques are assessed through simulation experiments and compared for several types of fault, including air-gap eccentricities, broken rotor bars and bearing damages.