Dispersion field analysis using discrete wavelet transform for inter-turn stator fault detection in induction motors

This paper presents an inter-turn short circuit detection through the measurement of external magnetic field of induction motor. The technique of analysis is based on discrete wavelet transform energy in transient state where the machine operates at start-up and standstill states. The wavelet offers a very fine analysis of the unidimensional and two-dimensional signals. Through multilevel discrete wavelet decomposition, it is possible to extract the information for the prognosis of the fault in variable state of the machine. The stored energy calculation in each decomposition can evaluate the severity of the fault compared with the analysis of only the external magnetic field.