Joint Estimation of Resistance and Speed With Auxiliary State of Active Flux for Sensorless IPMSM Drives

This article addresses the adaptive observer design problem for sensorless interior permanent magnetic synchronous motor (PMSM) drives with unknown stator resistance. When following the adaptive observer designs in the literature, there are difficulties encountered during tuning the observers' coefficients. Two aspects contribute to the complexity of tuning: the use of locally stable designs and the use of reduced-order models. Unfortunately, owing to the unknown regressors in the conventional interior PMSM (IPMSM) model (including the extended electromotive force model and active flux model), no full-order adaptive observer for sensorless IPMSM drives can be proved to be globally stable. To get rid of the unknown parameter regressors, a parameter-dependent auxiliary state is designed. The IPMSM model is transformed into a new coordinate, in which its dynamics show the characteristic of the Brunovsky form and the known regressors. With the proposed IPMSM model, a series of globally stable adaptive observer theories can be applied directly to the construction of full-order observers. Rigorous theoretical and experimental verifications reveal that a nonlinear parameterized high-gain observer provides superior dynamic performance and robustness.