Sensorless Control of Surface-Mounted Permanent Magnet Synchronous Motor Drives Using Nonlinear Optimization

This paper proposes a novel nonlinear optimization-based sensorless control scheme for surface-mounted permanent magnet synchronous motors (SPMSM) over wide speed range operations. Two algebraic constraints derived from the voltage equations and the flux equations of the SPMSM system are fundamental to the establishment of the proposed scheme. The position estimation algorithm generates the rotor position estimate via the stator flux estimates obtained by minimizing an equality constrained scalar function. Specifically, the flux information is extracted both from the voltage-based flux model and the current-based flux model, which contributes to its robustness against the inherent weakness of the pure integrator to dc offsets. Meanwhile, a nonlinear observer, based on the standard gradient search, is constructed for speed estimation. The proposed method is unique because the position estimation and the speed estimation are achieved independently; thus, the complexity associated with the speed dependence in position estimation is eliminated. Further, it is simple to implement. The feasibility and effectiveness of the proposed position algorithm and the speed observer have been verified by experiments under different operating conditions.