Three-stage model predictive control for modular multilevel converters with comprehensive performance optimization

Model predictive control (MPC) is recognized as an efficient control method for the modular multilevel converter (MMC), owing to its advantages, such as good robustness, rapid dynamic response, and multi-objective control. However, due to the coupling relationship between the ac-side current and the circulating current, the existing MPC has an impact on the ac-side current while suppressing the circulating current. In this paper, the relationship between the ac-side current performance and circulating current suppression is discussed in detail, and a three-stage MPC (TS-MPC) strategy is proposed to optimize the comprehensive performance. With the ac-side current control, circulating current control, and comprehensive optimization control, the optimum performance of both the ac-side current and circulating current suppression is realized while maintaining a low computational burden. Moreover, a grouping sorting algorithm is designed to reduce the calculation burden and to balance the capacitor voltages. The steady-state and transient performances of the proposed TS-MPC strategy have been verified by experimental results, which validates its correctness and effectiveness.