Sampled-data observer design with time-varying scalar gain for sensorless control of induction motors

Abstract

ABSTRACTA sampled-data sensorless observer is proposed for induction motors. A classical high-gain observer is associated with a time-varying gain between two sampling instants. In this paper, we show the interest of the use of a specific time-varying gain for sampled-data observers. This observer especially has to estimate the rotor speed which is the signal to be controlled, and also the flux norm and the load torque when only the stator currents are measured. We show that through some simulations, the good choice of a design parameter allows to improve the performances of the observer or to increase the bound of sampling interval.