The integration of machine fault detection into an indirect field oriented induction machine drive control scheme-the Vienna Monitoring Method

This paper proposes a new approach for an on-line induction machine monitoring scheme tailored for variable speed drives. As the application of the commonly employed spectrum analysis techniques is complicated by the inverter drives inherent frequency variation, the Vienna Monitoring Method avoids any frequency analysis and observes instead the machine state with the help of on-line models. The key for the fault detection is thereby the comparison of the estimated machine states out of two independent machine models with different model structures. While the models respond identically to the regular machine operation they diverge in case of a structural machine asymmetry. Experimental results from an IGBT drive show that a single rotor bar resistance change can be detected unequivocally. The method is thus sensitive enough to detect an upcoming fault in a very early state. As the extent of the cage asymmetry is estimated too, maintenance or repair can be conveniently scheduled on time.