Discrete-time backstepping induction motor control using a sensorless recurrent neural observer

This paper deals with the problem of controlling the discrete-time induction motor model based on a sensorless observer with only currents measurements. First a recurrent high order neural observer for the unknown plant is designed, then a high order neural network is used to emulate a control law designed by the backstepping technique. The learning algorithm for both neural networks is based on an extended Kalman filter. The applicability of the proposed observer-controller scheme is tested via simulation.