Model-Free Predictive Current Control for Dual Three-Phase PMSM Drives With an Optimal Modulation Pattern

This article presents a model-free predictive current control (MFPCC) method for dual three-phase permanent magnet synchronous machines with an optimal modulation pattern. First, an ultralocal model based on the
dq
subspace is established, and a novel nonlinear disturbance observer is designed to effectively compensate for voltage distortion caused by inverter characteristics. Additionally, the introduction of an intermediate variable eliminates the estimation peaks in the high-frequency state, improving both steady-state and transient performance of the controller while suppressing the influence of parameter mismatch on control performance. Second, two candidate modulation patterns and their corresponding pulse synthesis patterns are predefined, and the closed-loop control of the
xy
subspace is realized through an efficient cost function. The switching mode with the best
xy
current performance at the current moment is then switched in real time to obtain improved steady-state performance and deeper harmonic current suppression. Finally, the experimental results have verified the effectiveness of the proposed MFPCC method.