Soft starting of induction motors using neuro fuzzy and soft computing

Soft starters of induction motors are used in large number of applications such as blowers, fans, mixers, crushers, grinders, pumps and many other modern industrial applications. Voltage controller is the basic part of soft starter which is controlled to adjust the inrush current and developed torque. This paper presents a soft starter which is based upon artificial neural networks (ANN) and adaptive neuro fuzzy inference system (ANFIS). The neural network implements the feed back estimator while ANFIS with the help of ANN estimator adjusts the firing angle of thyristors of AC voltage controller under different loading conditions. The presented implementation gives satisfactory and promising results as compared with DSP-based controller. The presented approach can be used with off-line training as well as with on-line training and hence can solve the problem of on-line computation of firing angle. The paper also presents the comparison of two neural networks for the performance of soft starter.