Model-Free Predictive Current Control for a Nine-Phase Open-End Winding PMSM With Synthetic Virtual Voltage Vector

Model-free predictive current control (MFPCC) using an extended state observer (ESO) does not depend on the accurate motor model, exhibiting strong robustness. Using the virtual voltage vector (V<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup>) with zero harmonic plane voltage components as the control set effectively improves the computational efficiency of the finite set model predictive control for nine-phase open-end winding permanent magnet synchronous motor (OWPMSM). However, the MFPCC method using V<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> has encountered the problem of deteriorated steady-state performance. To solve this problem, this article establishes the ultra-local models and the ESOs for each subspace of the nine-phase OWPMSM, and the influence of high-frequency interference on MFPCC is analyzed. The utilization of synthetic V<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> along with a novel model-free duty ratio optimization method is proposed to generate the reference voltage vector and suppress high-frequency disturbances. Additionally, a bandwidth tuning method for the ESO that considers high-frequency disturbances is proposed to achieve optimal control performance. The experimental results have verified the effectiveness of the proposed MFPCC with synthetic V<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> method.