Sensorless FCS-MPCC PMSM Drives with Improved Sliding Mode Observer and Low-Complexity Discrete Vector Selection - An Assessment

This paper proposes a virtual voltage vector-based finite-control-set model predictive control (FCS-MPCC) for a sensorless permanent magnet synchronous motor (PMSM) drive. First, an adaptive sliding-mode back electromotive force observer is integrated to improve the observer accuracy, especially in terms of rotor field speed and position. Then, a simple virtual voltage vector selection algorithm is proposed to increase the input voltage vector resolution of FCS-MPCC while keeping the real-time computation burden low. The viability and performance of the proposed FCS-MPCC the sigmoid function-based sliding mode observer and the low-complexity virtual vector algorithm is assessed comparatively against the classical actual voltage vector-based FCS-MPCC with and without the improved sliding mode observer.