Model predictive control of modular multilevel converters

With the attractive features, modular multilevel converters are widely utilized in the area of high-voltage and large-capacity applications. The discrete state equation was derived, as the modular multilevel converter (MMC) was a multi-input multi-output system, a simplified finite control set model predictive control strategy was proposed, which could reduce the computing cycles and predictive value calculating amounts effectively. Optimal output voltage level calculations and modified capacitor voltage balancing control were included in the proposed control strategy. Firstly, an output voltage level was solved, based on ac-side current and circulating current equations; with this voltage level and its nearest levels, the finite control set was built. Considering the control of summation and difference of capacitor voltages in upper and low arms, the optimal output level was solved. A modified voltage balancing control method was also presented. An addition switch action was introduced into sorting algorithm as a voltage increment; in this way, voltage balancing control and switching frequency reduction were achieved effectively. Performances of the proposed control strategy were evaluated based on the time-domain simulation model in PSCAD/EMTDC; the simulation results verified the effectiveness of the proposed control strategy. © 2016 Chin. Soc. for Elec. Eng.