Model-free predictive current control of IPMSMs based on an improved extended state observer with parallel resonant controller switching control strategy

Abstract

Conventional deadbeat predictive current control (DPCC) relies on an accurate model, making it unable to handle both DC and AC unmodeled disturbances present in interior permanent magnet synchronous motor (IPMSM) drive systems. Model-free predictive current control (MFPCC) based on an extended state observer (ESO) with a ultra-local model reduces sensitivity to certain motor parameters and enhances system robustness. Nevertheless, this approach is limited by the bandwidth of the ESO, leading to poor disturbance observation for AC disturbances under multiple operating conditions, resulting in non-negligible current harmonics in the system. This paper introduces an enhanced MFPCC based on an improved extended state observer with parallel resonant controller switching control strategy(PRCSC-IESO). It observes major AC disturbances without amplitude attenuation and phase delay. Additionally, the introduction of a vector resonant controller (VRC) further suppresses current harmonics. The proposed observer can adaptively track AC disturbance frequencies and switch the number of VRCs according to actual rotational speed. The proposed MFPCC eliminates both DC and AC disturbances without requiring additional parameter information. Finally, experiments on a 7.1kw IPMSM validate the effectiveness of this approach in suppressing total harmonic distortion (THD) across various speed ranges and load conditions. Experimental results demonstrate that, compared to the traditional ESO method, the proposed method improves harmonic suppression capability by more than 40%. Additionally, the introduction of the switching control strategy reduces execution time by 24.4% at the rated motor speed, thereby enhancing the system’s operational efficiency.