Rotor resistance identification using artificial neural networks for a speed sensorless vector controlled induction motor drive

This paper presents a new method of rotor time constant estimation using artificial neural networks, for the speed sensorless implementation of the indirect vector controlled induction motor drive. The back propagation neural network technique is used for the real time adaptive estimation. The error between the desired state variable of an induction motor and the actual state variable of a neural model is back propagated to adjust the weights of the neural model, so that the actual state variable tracks the desired value. The performance of the neural network based estimator is investigated with the help of simulations and experiments, for variations in the rotor resistance from their nominal values, with both speed and load torque disturbances. A Programmable Cascaded Low-Pass Filter is used for the estimation of rotor flux, from the measured stator voltages and currents. The rotor speed is estimated from the flux angles and the estimated slip speed.