Model-Free Predictive Current Control With Virtual Harmonic Back-EMF Identification for Dual Three-Phase PMSM Drives

This article proposes an improved model-free predictive current control (MFPCC) scheme for dual three-phase permanent-magnet synchronous motor (DTP-PMSM) drives, in which virtual harmonic back electromotive force (back-EMF) identification is incorporated to achieve effective harmonic suppression. First, a comprehensive harmonic modeling framework is developed, showing that nonsinusoidal back-EMFs and voltage-source inverter (VSI) nonlinearities give rise to 12th- and 6th-order harmonic components in the dq and dqz subspaces, respectively. Second, to address these harmonic effects, a virtual harmonic back-EMF representation is introduced to unify VSI-induced distortions and spatial harmonic back-EMFs into an equivalent disturbance model. Based on this model, a current-mapping principle is developed to estimate the amplitude and phase deviations of the virtual harmonic back-EMFs. Third, the identified virtual harmonic back-EMFs are incorporated as a compensation term into the ULM voltage input, and the resulting augmented input is processed by the linear extended state observer (LESO)-based prediction to suppress harmonic disturbances beyond the LESO bandwidth. This mechanism enables robust harmonic suppression without complicated algorithms or extensive parameter tuning. Finally, experimental results validate the effectiveness of the proposed method over a wide speed range.