Kalman filtering applied to induction motors: A deterministic viewpoint

Abstract

This paper considers the use of the discrete time Kalman filter in the problem of flux estimation in induction motor drives. The Kalman filter is discussed from a deterministic point of view rather than the usual stochastic point of view. An application is considered in which a steady state and time variable version of the Kalman filter is applied to estimate rotor flux in an induction machine. In this application, the motor speed is assumed to be measured and the steady state Kalman filter gain matrices are gain scheduled with respect to the motor speed parameter. Also discussed is the relationship between the Kalman filter flux estimator and the full order Luenberger flux estimator.