Alternative Solution Regarding Problems of Adaptive Observer Compensating Parameters Uncertainties for Sensorless Induction Motor Drives

摘要

This article first shows that the advanced high-gain adaptive observer (AO) is able to estimate both speed and resistances during speed transients. However, it is also found that high-gain AO-based resistances estimation begins to diverge as the operating speed becomes higher. The problem is due to the fact that the model sensitivity of speed is much higher than that of resistances at high speeds. As a result, this article claims that the high-gain AO is not generally a good option to compensate parameters uncertainties in sensorless induction motor drives for all working conditions. As an alternative, this article proposes a scheme that not only has the potential of high dynamic performance but also avoids the aforementioned "sensitivity problem." Besides, interesting topics, such as speed-dependent resistance estimation, sensorless inertia identification, slow speed reversal test, and reduced-order natural speed observer, are all covered in this article. Effective experiment verifies the feasibility of the proposed scheme, which is deemed to have potentials for practical applications.