Neural network based input output feedback control of induction motor

Abstract

The present paper gives a new neural network based adaptive control for the induction motor using input-output feedback linearization control. A simple multilayer neural network is used for the estimation of rotor and stator time constant inverses as well as the load torque. The suggested estimator is a model reference adaptive control based which uses the measured and estimated motor states such as the currents and the speed to generate an online learning algorithm for the neural network parameters.