Predictive Controller Based on Switching State Grouping for a Modular Multilevel Converter With Reduced Computational Time

Recently, a modular multilevel converter (MMC) has become one of the most popular converters in medium/high-power applications. In this paper, a new technique is presented to decrease the computational time of the MMC by using switching state grouping for the predictive controller. The proposed technique has the ability to control the output current, minimize the circulating currents, and balance all capacitor voltages, simultaneously. For a predefined number of MMC submodules, the number available switching states is known. The proposed technique uses two stages of predictive modules. In the first stage, the switching states of the MMC are divided equally into a number of groups, controlled simultaneously by the same number of predictive modules. The optimal number of the switching state groups in the first stage is obtained. In the second stage, only one predictive module is used to obtain the optimal switching state of the MMC without any other complexities or additional sorting controllers. The effectiveness of the proposed control technique has been verified by using MATLAB/SIMULINK software considering steady state and dynamic analysis. The results of the proposed technique are compared with the conventional one in terms of computation time, verifying better performance of the proposed method.