Two-step Model-free Predictive Current Control for PMSM

One-step and two-step model-free model predictive current control (MFPCC) for permanent magnet synchronous motor (PMSM) is established using ultra-local model and algebraic parameter identification technique is adopted to determine uncertain terms of first-order and second-order ultra-local. The effect of window sequence length of ultra-local on control performances is analyzed. And robustness for parameters of MFPCC is studied. Real-time performances of conventional model predictive current control (MPCC) and MFPCC are tested based on STM32H743 micro control unit platform. Simulation and experimental results testify the effectiveness of one-step and two-step MFPCC for PMSM. With the increase of window sequence length of the first-order ultra-local, the control performance improves and tends to saturate, but window sequence length of the second-order ultra-local has less effect on control performance. One-step and two-step MFPCC for PMSM has strong robustness for the change of motor parameters. With the increase of window sequence length, calculate burden of MFPCC increases slightly which has less effect on real-time performance. Real-time performance of one-step MPCC and MFPCC are almost the same and real-time performance of two-step MFPCC is better than two-step MPCC.