Stator resistance estimation for MRAS-based speed sensorless vector-controlled switched reluctance motor drive

Abstract

In this paper, unipolar excitation-based speed sensorless vector-controlled switched reluctance motor (SRM) drive is addressed. The use of position/speed sensors has few apprehensions in terms of price, dependability, compactness, especially from electric vehicle (EV) application view point. Therefore, for speed/torque control of SRM, it is mandatory to have some estimation technique which estimates the speed/position satisfactorily and with ease. In this paper, F-MRAS (flux-based model reference adaptive system) speed estimator is employed for estimating the speed/position. The reason for selection of this speed estimator is because of its simplicity, no large lookup table requirement and no external circuitry requirement for pulses injection. This estimator depends on stator resistance of the machine. Therefore, to make the speed estimator robust to stator resistance variations, on-line estimation of the resistance and its compensation is required. From this view point, a novel stator resistance estimator is proposed which estimates the resistance on-line. Further, this paper discusses in detail about the stability analysis of proposed resistance estimator and F-MRAS-based speed estimator. The speed estimator in combination with the proposed resistance estimator performs well in both motoring and regeneration modes. An exhaustive simulation of the drive has been carried out in MATLAB/SIMULINK. Also, the results are confirmed experimentally on a dSPACE-1104 platform.