Automatic fault classification in power electronics using time–frequency distribution and neural networks

Abstract — A new method of fault analysis and detection by signal classification in frequency converters is presented. The Wigner-Ville time frequency distribution is used to produce the representation of the signal and the probabilistic neural network as a classifier. The accuracy and robustness of the proposed method is investigated on signals obtained during the different fault mode operations of the industrial frequency converter. Index Terms — Fault diagnosis, Neural Networks applications, Power system faults, Pulse width modulated power converters