A Robust Parallel Predictive Torque Control Based on a Proportion Integral Cost Function for PMSM

The control performance of conventional model predictive torque control (MPTC) is restricted by two key factors: motor parameter mismatches and weighting factors tunning. To solve the above problems, a robust parallel model predictive torque control (RP-MPTC) is proposed in this paper. Firstly, a cost function in the form of proportion integral is designed to improve robustness. Based on the proposed cost function, not only the current moment error is considered, but also the error of the past moment is accumulated. In this way, the strategy can improve the robustness of system parameters without disturbance observers. Then through a parallel structure, this structure simultaneously optimizes the torque and flux and then selects the common optimal voltage vector to eliminate the weighting factor. The contributions of the proposed RP-MPTC strategy are that it can improve the robustness of parameter mismatches and eliminate the weighting factor to avoid cumbersome the weighting factor tunning. Finally, the experiment is carried on 1.5kW permanent magnet synchronous motor (PMSM) to verify the superiority of RP-MPTC.