A Neuro Fuzzy Application: Soft Starting of Induction Motors with Reduced Energy Losses

Abstract

Soft starters are used in electrical drives for the smooth starting of blowers, fans, mixers, crushers, grinders, and pumps, as well as in many other modern industrial applications. This article presents a soft starter that reduces energy losses during the start-up process of an induction motor. A sensor-less technique has been used to enhance the response of the system. The artificial neural networks, used in the soft starter, have been compared for the estimation of different parameters. The adaptive neuro fuzzy inference system has been developed to control the speed and torque of the motor with the constraint of “reduction in energy losses” during the start-up process. A neural network implements the feedback estimator, thus eliminating the need for slow mechanical sensors, while the adaptive neuro fuzzy inference system with the help of artificial neural network estimators adjusts the firing angles of the thyristors (of an AC voltage controller) under different loading conditions. The control system has been implemented using a TMS320F2812 (Texas Instruments, Texas, USA) processor. The presented approach can be employed for both the off-line and on-line trainings and, hence, can solve the problem of on-line computation of firing angles.