Neural Networks and Neuro-fuzzy Based States and Parameters Estimation in Induction Motor Sensorless Drive

During the last decade, speed sensorless field-oriented control of induction motor has given a particular attention by researchers worldwide and a great number of papers have been published on this issue. In most of them, the authors proposed the speed estimation algorithms based on Kalman filter theory, neural networks and model of reference. In indirect vector control strategy, the accurate knowledge of the rotor resistance is critical to ensure field-orientation. However, very few papers have been published on the simultaneous estimation of the speed and the rotor resistance. This paper describes the use of artificial neural networks and neuro-fuzzy networks for the simultaneous estimation of the speed, rotor flux and rotor resistance of an induction motor. This achievement is in authors’ opinion a great contribution. Simulation results showed the effectiveness of the proposed schemes.