Adaptive Back-EMF Observer Based on Filtered Transformation for Sensorless SPMSM Control

Most existing back-electromotive-force (EMF) observer necessitates a phase-locked loop (PLL) for real-time observer parameter updating, i.e. the rotor's angular speed. However, the observer-PLL coupling effect complicates the back-EMF observer design and poses challenges in ensuring stability. In this paper, a filtered transformation-based adaptive back-EMF observer (FT-AO) for surface-mounted permanent magnet synchronous motors (SPMSMs) is presented. The FT-AO enables direct estimation of back-EMF without relying on a PLL, avoiding the observer-PLL coupling and thus simplifying the control parameter design. Additionally, the asymptotic stability of the FT-AO is rigorously proven using Lyapunov's theorem. Experimental results confirm the superiority of the FT-AO in terms of dynamic/steady-state performance.