On-line Estimation of Magnetizing Inductance and Rotor Resistance in Extended Kalman-Filter for Induction Machines

This paper proposes an on-time estimation of induction machine parameters based on the extended Kalman-filter. So far, the real time performance of Extended Kalman-filter algorithms has not been validated in the variation of motor parameters. Furthermore, the conventional parameter estimations in extended Kalman-filter has not been developed accurately based on the correct non-linear state space model. This paper proposes a new state space model of induction machine included motor parameters with less dependency to other model variables. This leads to achieve more accurate estimation results even in situation where a sudden variation of motor parameters happens. This paper describes the proposed method analytically. Simulation results for an induction machine are presented and discussed.