Skip to Main Content
This paper presents a new approach to Fault detection and diagnosis in power system. Discrete wavelet transformations (DWT) combined with neural networks (NN) have been applied to a typical three phase inverter. A set of faults have been examined, such as inverter IGBT open-circuit fault, leg open fault. The input signals of this algorithm are the three-phase stator currents. Identification and classification uses approximation and details at levels 6 of these currents. The results of simulation show that the proposed technique can accurately detect identify and classify effectively the faults of interest in the power system.
Industrial Electronics, 2006 IEEE International Symposium on (Volume:2 )
Date of Conference: 9-13 July 2006