Stator current signature analysis for fault diagnosis in permanent magnet synchronous wind generators

Abstract

Inter-turn short circuit faults are the most common stator winding faults in permanent magnet wind generators. Their early detection is vital to avoid subsequent damages which can led to catastrophic machine failures. To accomplish this task stator current signature analysis has been recently proposed for identifying inter-turn faults via a non-invasive, non-destructive and cost effective approach. Since no best stator current signature analysis approach is recognized as optimal, the paper takes in consideration the one able to offer the minimum severity degree which is Hilbert Huang Transform. To propose some improvements of the latter in stator current signature analysis, a real 60 kW-25 rpm permanent magnet wind generator is firstly designed via finite element based approach in FLUX® environment. Then SCSA in faulty conditions is performed via proprietary software highlighting how the introduced modifications could further enhance the signal processing algorithm performance.