A position-sensorless vector control for doubly-fed induction machines using adaptive reduced-order observers

In this paper, a model-base approach is employed for the rotor-position estimations of doubly-fed induction machines. Two types of adaptive reduced-order observers; based on stator and rotor models, are conducted in order to estimate the rotor position and rotor speed simultaneously. The proposed estimations schemes are straightforward and simplicity for implementation regardless of the knowledge of the stator flux which is usually approximated through the low-pass filter. The proposed rotor-position estimations are integrated with the stator-flux vector control. Experimental results of various operating conditions are provided in order to verify the proposed theoretical concepts. The proposed sensorless system can effectively operate in both motor and generator regions, and the performances in synchronous, sub-synchronous, synchronous and super-synchronous modes are additionally evaluated.