Ultra-Local Model-Free Predictive Control with Switching Strategy of PMSMs for Two-Phase Stationary Current

This paper presents a high-performance switching ultra-local model-free control strategy for permanent magnet synchronous motors (PMSMs) in two-phase stationary coordination. First, an ultra-local model-free predictive control scheme that incorporates a tracking error term is proposes in this paper, allowing for enhanced control performance with reduced parameter tuning. Second, the exponential term is introduced into this strategy to enable a more accurate fitting of motor current response, thereby enhancing the system's dynamic performance. Finally, simulation results show that the proposed switching model-free predictive control (PSMFPC) achieves the shortest rise time during no-load startup. The d-axis and q-axis current oscillations of PSMFPC are significantly reduced by 26.79% and 15.64% respectively. Inductance mismatch has a significant negative impact on traditional PI and model-free predictive control (MFPC) strategies, nearly doubling the dq-axis current oscillations, while PSMFPC demonstrates higher robustness.