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

In this work are developed two on-line identification methods for induction motors based on discrete models obtained when we consider continuous systems excited by sinusoidal signals. The first method uses the discrete model of the homopolar machine in the stationary frame to estimate the stator resistance and the stator leakage inductance and the discrete model of the linear system existent among the stator flux and stator current in the rotor frame to estimate all the electric parameters of the motor. The second method, besides this last model, presupposes to estimate the stator resistance with DC excitation added to the supply voltage of the motor and the knowledge of the motor class, in order to estimate the other electric parameters through classical methods of least squares parameters estimation. Simulation and experimental results illustrate the proposed methods.