NEURAL NETWORK SPEED CONTROLLER FOR INDUCTION MOTOR BASED ON FEEDBACK LINEARIZATION

A simple structure neural network (NN) is presented to control the speed of the induction motor in this paper. The scheme consists of a neural network controller, a reference model, and an algorithm for changing the NN weights in order that the speed of the drive can track of the reference speed. The rotor flux is estimated by the simplified rotor flux observer on the rotor reference frame and the feedback linearization theory is used to decouple the rotor speed and the flux amplitude. The proposed controller, without using state space dynamics and by employing only the input-output information can produce the control input for motor. However, the NN controller is robust to the bounded parameter variations and external disturbances. A qualified speed tracking and load regulating responses can be obtained by the proposed controller.