A novel Model-Free predictive current control for PMSM with Stagnation-Free current difference update scheme

Model predictive current control (MPCC) for permanent magnet synchronous motors (PMSMs) relies heavily on the accuracy of the system model and parameters. To enhance parameter robustness, model-free predictive current control (MFPCC) has been developed, which depends solely on current sampling without the need for motor parameters. Conventional MFPCC stores current difference information in a lookup table (LUT), but only one LUT element can be updated per control period, leaving the remaining entries stagnant. The stagnation of current difference updating (SCDU) is the core issue of MFPCC. It degrades the reliability of current differences, further deteriorating control performance and increasing current ripple and harmonics. Thus, this paper proposes a novel MFPCC with stagnation-free current difference update scheme to overcome SCDU and improve performance. First, an effective and fast calculation of inductances is developed using historic sampling information. Then, a dynamic update mechanism of current difference is presented. This mechanism ensures complete SCDU elimination by updating all current differences within each control period, thus guaranteeing the reliability of current differences. Moreover, the proposed method is independent of any initial parameters with excellent parameter robustness. Finally, the proposed method is compared with conventional MPCC and previous MFPCC methods. The superiority of the proposed method is experimentally verified in an interior PMSM.