23 December 2024 in IEEE Transactions on Power Electronics
This article proposes a sensorless control method for the permanent magnet synchronous motors (PMSM) combining the fifth-order generalized integral flux observer (FOGIFO) and the extended state observer-based phase-locked loop (ESO_PLL). First, due to various nonideal factors, including parameter mismatch, sampling errors, and inverter nonlinearity, the rotor flux estimated by traditional flux observers suffers the dc offset and harmonics. To solve these issues, a FOGIFO incorporating a frequency-locked loop (FLL) is proposed, and the FLL ensures that FOGIFO dynamically tracks the frequency of the back electromotive force signal. This guarantees superior integral characteristics at the central frequency point. Meanwhile, the FOGIFO pole distribution is optimized to enhance the harmonic suppression performance. Subsequently, the analysis of the conventional phase-locked loop is carried out in terms of the error transfer function. The ESO_PLL is used for rotor position and speed observation, which significantly improves the dynamic tracking performance. Moreover, since the proposed method is based on a voltage model in stationary coordinate system, it is applicable to both surface and interior PMSM, and simulations and experiments have verified its effectiveness.
22 July 2025 in IEEE Transactions on Power Electronics
This article presents a robust sensorless control strategy for surface-mounted permanent magnet synchronous motors. The approach integrates an enhanced full-order supertwisting sliding mode observer (STA-FSMO) with a feedforward phase-locked loop utilizing a harmonic-resonant proportional–integral (HR-PI) controller featuring a tunable forward gain. Conventional sliding mode observers often exhibit high-frequency chattering and limited resilience against linearly growing perturbations, degrading estimation accuracy and convergence. To address these limitations, the proposed STA-FSMO employs a dual-layer correction mechanism: linear feedback ensures rapid response during transients, while a nonlinear supertwisting action guarantees smooth, finite-time convergence near equilibrium. This design effectively suppresses linearly growing disturbances and ensures robust estimation under dynamic conditions. Rigorous Lyapunov-based analysis confirms global stability, finite-time convergence, and robustness against parameter uncertainties. The HR-PI controller mitigates inverter-induced back-EMF harmonics, particularly the fifth and seventh components, while its tunable forward gain quantitatively enhances dynamic performance metrics, such as overshoot and rise time. Parameter sensitivity analysis verifies bounded estimation errors under stator resistance and inductance variations. Extensive experimental results validate the method’s superior accuracy, robustness, and practical feasibility, including critical low-speed and zero-to-reverse loaded operations.