Feedback linearization based sensorless direct torque control using stator flux MRAS-sliding mode observer for induction motor drive

International audienceThe high-performance Direct Torque Control (DTC) requires accurate knowledge of flux and speed information. Furthermore, the elimination of sensors leads to reduced overall cost and size of the electric drive system and subsequently improving its reliability. This paper proposes an effective sensorless direct torque control scheme for induction motor drive. The proposed scheme consists of enhancing the decoupling structure and variable estimation as well. Therefore, an enhanced direct flux and torque control based on feedback linearization is implemented in one hand. This allows obtaining a linear decoupled control together with minimized flux and torque ripples. In another hand, a combined sliding mode observer and model reference adaptive system is associated with the control scheme as sensorless algorithms for rotor speed and flux estimation. This conjunction is intended to enhance the sliding mode observer performances especially at low speed operations and reduce its sensitivity to noise and system uncertainties as well. The effectiveness of the proposed control algorithm has been verified through simulation and experimental work using MATLAB/Simulink software and dSpace 1104 implementation board respectively