Sliding-Mode MRAS Speed Estimator for Sensorless Vector Control of Double Stator Induction Motor

The weakness of the Direct Vector Control (DVC) is lack of an estimation of the flux amplitude and the rotor position with high accuracy. These quantities are sensitive to parameter variations; it is important to use a robust estimation system for estimating the rotor flux with respect to parametric uncertainties. In this paper the sliding mode speed sensorless vector control based on the Model Reference Adaptive System (MRAS) of double stator induction motor is presented. First, the models of the double stator induction motor and the DVC are proposed. Second, the MRAS technique of the DSIM is adopted. In order to ensure a robust sensorless control, the sliding mode technique for the estimation system was used. The results showed the presented estimator has a positive effect on the system behavior especially in changing the reference and/or the parameters variation.