Detection of Broken Rotor Bars in Induction Motor Using Derivative Free Kalman Filters

This paper deals with design and implementation of Joint Unscented Kalman filter (JUKF) and Dual Unscented Kalman filter (DUKF) for the detection and monitoring of rotor bar faults in induction motor under simulation studies. A broken rotor bar essentially leads to an increase in rotor resistance of the induction motor. The methodology used is basically model based fault detection in which the problem is treated as one of detection and estimation of parameter variation. An extensive monte carlo simulation study has been carried out to assess the relative performance of the two filters under various operating conditions. The results of the simulation studies show that DUKF is more sensitive to rotor resistance variation over wide range of tuning parameters and gives better performance than JUKF in detecting and estimating the rotor resistance . However DUKF also shows high sensitivity towards load disturbances.