Model-based broken rotor bar detection on an IFOC driven squirrel cage induction motor

In this paper, we propose an on-line monitoring scheme to detect broken rotor bars on IFOC driven squirrel cage induction motors. The drive's variable speed nature complicates the use of classical spectrum analysis techniques. The proposed model-based monitoring scheme does not rely on spectral methods; instead, it monitors, a carefully selected induction motor state, using an on-line observer. The key to fault detection is the development of a simplified dynamic model of a squirrel cage induction motor with broken rotor bars, and the selection, based on techniques from differential geometric theory, of the induction motor state to monitor. Numerical simulations of very different motors validate the model and the monitoring scheme.