Stability Improvement of Adaptive Full-Order Observer for Sensorless Induction Motor Drives in Low-Speed Regenerating Mode
Hongwu Chen, Jian Li, Yang Lu, Kai Yang, Linghao Wu, Liu Zhi
Abstract
In the applications of high power induction motor railway traction systems, the switching frequency of power devices is relatively low. This restriction will severely reduce the accuracy and stability of conventional discrete model for adaptive full-order observer (AFO) in extremely high-speed region. What’s worse, due to the neglect of flux error during speed estimation derivation, the conventional AFO cannot maintain stability under low-speed regenerating mode. The above two reasons limit the application of AFO in rail traction systems for full speed range. To address this problem, this paper utilizes a high-accuracy discretization which is achieved in combined reference frames and feedback gains are designed in z-domain to ensure stability and dynamic performance of AFO. Moreover, a modified speed estimation mechanism with the error of d-axis current is developed to guarantee stability in extremely low-speed regenerating mode. Rigorous simulations and experiments are performed to verify that the proposed AFO model achieves excellent sensorless control performance in full speed range.