Fault detection techniques using current signature analysis methods

There are many condition monitoring methods such as vibration monitoring, thermal monitoring, chemical monitoring and acoustic emission monitoring. But all of these monitoring methods require expensive sensors and specialized tools. However, the condition monitoring method and fault diagnosis based on motor current signature are a better option since they do not require additional sensors. In this research, a novel criterion function of wavelet processing signal is introduced to diagnose the broken rotor bars in three-phase squirrel cage induction motors. This criterion function facilitates the precise diagnosis of the faults in induction motors under load variations. It uses wavelet transforms available in LabView software to process the stator current signals in the faulty induction motors to extract the wavelet coefficients in a specific time-frequency bands. Furthermore, spectrum analysis of the stator currents around the fundamental frequency is used to diagnose the faults. It is shown that the amplitudes of the frequency harmonics components fb=fs(12s) are influenced by the number of broken rotor bars, the exact location of broken rotor bars and the motor loading condition.