Detection of unbalance in a wind turbine by using wavelet packet transform and vibration signals

Abstract

Wind turbines (WTs) are increasingly used in many countries for clean and green electric generation. Condition monitoring and fault detection of WTs reduce both downtimes and costs in the electric service. In this regard, it is important to ensure their safety and reliability. This paper presents a methodology based on the wavelet packet transform (WPT) for detection of unbalance fault in a WT. In general, the methodology consists of the acquisition and analysis of vibration signals coming from the WT. For vibration signals, WPT is firstly applied. Then, one node of the wavelet packet tree is analyzed using an energy index. This index is computed as a fault feature. Finally, a statistical analysis is carried out in order to observe the capability of discriminating between a nominal condition and a fault condition. Obtained results show that the proposal can detect the unbalance fault.