Sensorless Speed Estimation for Induction Motor based on Neural Network

Abstract

Method of sensorless speed estimation for induction motor based on neural network has been studied. With stator currents as input and the motor speed as output, a neural network has been trained by the principle of training off-line, estimating on-line. The neural network can imitate the activity of the induction motor well and estimate the motor speed on-line. The dynamic performance of the method for induction motor fed by the pure sine wave supply or inverter was focused on. And average errors were used as a measure for evaluation of the method. Experiments have been implemented and the effectiveness of the method used was validated