Sensorless Stator Field Oriented Direct Torque Control with SVM for Induction Motor Based on MRAS

This paper deals with improvement of Direct Torque Control strategy for induction motor (IM) drive. Since the main disadvantages of the classical DTC are high torque/flux ripples and current distortion, this paper inserts the space vector modulation in order to reduce the ripples by maintaining a constant switching frequency. Besides, the fuzzy logic controllers will replace the traditional proportional-integral (PI) controllers for stator flux and torque regulation and to ensure an accurate reference tracking and a robust response against different uncertainties such as external disturbance and parameters variation. Furthermore, a stator flux based Model Reference Adaptive System (SF-MRAS) is designed as a sensorless algorithm for the estimation of rotor speed. This estimator can improve the performance of the controlled system by increasing its reliability and decreasing the cost of the speed sensor. The global control algorithm has been investigated via numerical simulation and real-time experimentation using Matlab/Simulink with dSpace 1104 signal card.