Dynamic performance limitations for MRAS based sensorless induction motor drives. Part 2: Online parameter tuning and dynamic performance studies

The paper investigates the dynamic performance limits of a sensorless cage induction motor drive utilising an MRAS based speed and flux estimator operating within a direct rotor flux orientated (DRFO) vector controller. A 4 kW experimental sensorless drive rig has been constructed employing new on line tuning techniques for the rotor time constant and stator resistance. These techniques, together with their implementation and experimental performance, are described. The tuning of the rotor time constant allows for robust speed estimate accuracies of 0.1 rev/min to be attained. The tuning of the stator resistance to near 1% allows for improved low speed performance and enhanced speed loop dynamics. The resulting speed loop dynamics of the tuned sensorless drive is compared experimentally with a sensored vector control drive and the limiting factor for both is seen to be the quality of the speed feedback signal. A quantitative goodness factor for comparing the speed loop dynamics is introduced and the comparison discussed.