A PMSG Wind Turbine Fault Diagnosis via Sparse Recovery Method

This paper presents a diagnosis method to detect, isolate and estimate faults occurring in permanent magnet synchronous generator (PMSG). This method is based on a dynamical sparse recovery (SR) algorithm that is able to reconstruct online and with finite-time convergence a sparse vector of numerous faults from few system measurements. The faulty PMSG system is affected by a partial inter-turn short circuit and voltage dip faults. Using Matlab/Simulink environment, some scenarios are presented to present the efficiency of SR method for diagnostic of PMSG system.