Double ESO Aided Nonlinear Flux Observer for Sensorless Control of PMSM

As one promising sensorless control algorithm, the nonlinear flux observer is widely used in industries. However, it is sensitive to parameter variations and the gradient search gain can greatly affect its stability. Therefore, this paper proposes an improved nonlinear flux observer, which corrects the stator flux estimation by adding an extended state observer (ESO) to reduce the parameter sensitivity. In addition, the position error compensation is realized by another ESO constructed in estimated rotation frame. The effectiveness of the proposed algorithm is verified experimentally.