An Efficient Dynamic Sequential Predictive Control of Module Multilevel Power Converters

Modular multilevel converters (MMC) have been widely used in high voltage direct current transmission. Model predictive control (MPC) has become a promising control alternative for MMC due to its good dynamic performance and multi-objective control ability. However, classical MPC requires traversing all switch vectors and causes huge computational burden. Therefore, the sequential optimization MPC is adopted to reduce the computational burden. But it has poor dynamic performance and control effect on suboptimal targets. This paper proposes a new efficient dynamic sequential MPC (DSMPC) method for MMCs. Two cost functions are designed forming a sequential structure, to achieve current tracking and circulating current suppression, respectively. The number of candidate combinations is automatically adjusted based on the operation mode, to ensure dynamic performance during transient state while alleviating the computational burden during steady state operation. The increased number of candidate combinations is related to the degree of control targets, which improves the effect of secondary control targets. Simulation and experimental results confirm that the proposed method reduces computational burden while maintaining superior steady and dynamic state performance.