Three-Vector-Based Smart Model Predictive Torque Control of Surface-Mounted Permanent Magnet Synchronous Motor Drives for Robotic System Based on Genetic Algorithm

Owing to their high performance and high-efficiency controllability, surface-mounted permanent magnet synchronous motors (SPMSMs) have been widely considered for various robotic systems. The conventional three-vector-based model predictive torque control (MPTC) is frequently applied to SPMSMs, while the adjustment of weight factors is difficult. Compared with the five-segment sequence output method, the three-segment sequence output method can effectively reduce the switching frequency. However, the three-segment sequence output method leads to large torque and stator flux ripple. For these issues, a three-vector-based smart MPTC method based on the optimal vector sequence optimized by a genetic algorithm is proposed. Firstly, the reference voltage vector output from the discrete-time sliding mode (DTSM) current controller is utilized to simplify the process of selecting the vectors, and it can enhance the robustness of the SPMSM system. Secondly, an improved cost function is employed to select the optimal vector sequence, aiming to minimize torque and flux ripple. Furthermore, the multi-objective genetic algorithm is leveraged to seek the Pareto solution for weight factors. As a final step, the efficacy of the designed MPTC approach is confirmed through simulations and experiments.