Detection of early-stage inter-turn short-circuits in an induction machine based on CWT of phase currents by means of deep neural network

Due to the dynamic technological development, increasingly sophisticated methods and algorithms are used in the field of diagnosis of electrical machines. Recently, ones of the most frequently proposed methods are based on advanced machine learning techniques. The article presents a method of analysis of the operating state of an induction motor based on continuous wavelet transform as well as deep neural network. The research focused on multiphase, symmetrical inter-turn short-circuits in the stator circuit.