Gain Compensation-Based Quasi-Resonant PLL With Adaptive Super-Twisting SMO for Position Error Suppression of IPMSM Sensorless Control

In this article, a gain compensation-based quasi-resonant phase-locked loop (GCQR-PLL) with adaptive super-twisting sliding-mode observer (AST-SMO) is proposed. The conventional SMO typically employs a low-pass filter (LPF) and a quadrature phase-locked loop (QPLL) for position extraction. However, the tracking performance of the conventional QPLL is unsatisfactory, leading to position estimation error under transient conditions. Furthermore, the LPF fails to mitigate the dc disturbance caused by detection errors in the estimated back-electromotive force (EMF), resulting in position estimation errors. To minimize the position estimation error, the proposed GCQR-PLL performs gain compensation under transient conditions, which ensures sufficient bandwidth for position extraction, and at the same time, it effectively suppresses the dc disturbance due to detection error in estimated back EMF by quasi-resonant (QR) structure. Additionally, the AST-SMO can diminish chattering at lower speeds while enhancing position estimation accuracy at higher speeds. The effectiveness of the proposed algorithm is verified under different operating conditions.