Abstract: Aiming at the issue of poor parameter robustness of the conventional three‍-‍vector model predic‍tive current control strategy for permanent magnet synchronous motor drive systems, an improved three-vector model-free predictive current control strategy based on time series is proposed to eliminate the effect of parameter mismatch and enhance the robustness of the system. To begin with, a time-series data‍-driven model is established to fit the input-output data as a discrete transfer function and combined with the recursive least squares method to estimate the model undetermined coefficients online and predict the required variables. In addition, the vector sector is reclassified to optimize the selection process of the three-vector combination. Moreover, the vector duty cycle direct calculation method is introduced to suppress the uncertainty effect of motor parameter ingress on the duty cycle calculation link, which further improves the system robustness. Finally, the simulation and experimental results show that the proposed strategy can ef‍fectively improve the model parameter robustness. The dq-axis current ripple is reduced, and the disturbances caused by the variation of motor parameters can be effectively suppressed.

Key words: permanent magnet synchronous motor（PMSM), model predictive control, model-free, time series, three-vector