NEURAL NETWORK BASED SPEED CONTROL OF INDUCTION MOTOR

Recently, the Neural Networks are widely used in power electronics and motion control systems. In this paper, the speed control of induction motor using Neural Network is presented. The high processing speed and control accuracy have been achieved by controlling of feedback signals in an induction motor drive. The controlling of feedback signals have been carried out by generating of pulse width modulated signals for the switching action of inverter. The pulse width modulation signals are generated by developed controller. The controller consists of Neural Network estimator and speed controller. The designed Neural Network estimator estimates flux, unit vector and torque by receiving machine terminal signals as input. The speed controller takes the actual speed from proximity sensor and set speed from speed setting unit into account and estimates the error between them. The proposed system has been compared with the developed Digital Signal Processor based controller and open loop system. From the comparison it is found that the constant speed has been achieved with variation of load. It is also observed that increase in efficiency and torque, decrease in regulation, faster execution speed and fault tolerance characteristics compared to Digital Signal Process based speed control.