On-line identification of induction motors through discrete models for sinusoidal signals

Abstract

This work presents a method for on-line identification of induction motors based on discrete model of the linear relationship existent among the stator flux and current in the rotor frame. The modeling arises by considering the sinusoidal nature of these signals, and their sampling without zero order holder. The method presupposes the estimate of these signals from their measurements in the stator frame, the estimate of the stator resistance with DC excitation, the measurements of the rotor angular velocity and the knowledge of the motor class, in order to one obtain the other electric parameters through classic methods of least squares. Simulation and experimental results illustrate the methodology.