Online detection for blade imbalance of doubly fed induction generator wind turbines based on stator current

Abstract

Stator current of the Double Fed Induction Generator (DFIG) is analyzed for blade imbalance detection of wind turbines (WTs). This paper firstly describes the fault electrical characteristics of DFIG in the blade imbalance scenarios. And the Hilbert demodulation method is presented for extracting fault frequency, also known as the shaft rotating frequency (1P). However, the operating surroundings of WTs change with time and disturbances easily appear near the impellor rotating frequency. Therefore, partial least squares (PLS) algorithm is introduced in order to eliminate especially the impacts of wind speed variation in consideration of the average value of absolute stator current and standard deviation of rotor speed. Finally, the experiments on commercial DFIG WTs have been taken in three operating scenarios. And the results show that stator current contains components at impeller rotating frequency. And comparisons are presented between the electrical detecting analysis and radial vibration test of main bearing. All these results, proved that the proposed method is effective for fault detection of blade imbalance for DFIG WTs.