The application of modern signal processing techniques to rotorfault detection and location within three phase induction motors

摘要

Previous work at The Robert Gordon University has shown thatfaults within the rotors of large three phase induction motors, such asbroken rotor bars, can be detected by monitoring and analysing the linecurrent taken by the machine during a no-load starting transient. Thisline current has been shown to contain frequency components which areindicative of these fault conditions. Under transient conditions thesecomponents are however nonstationary in both the time and frequencydomains. A variety of signal processing strategies applicable to timevariant data such as the spectrogram, The Wigner Ville distribution andwavelet decomposition have been implemented and their success indetecting these non-stationary components evaluated. The most suitableof these techniques has been used to determine the occurrence andseverity of motor faults. Preliminary work suggests that thesetechniques may also be used to detect frequency components indicative tothe location of the fault