Indirect Model Predictive Control Scheme with Highly-reduced Computational Burden for Three-phase MMC

Due to its several benefits like Modularity, scalability and improved reliability, the Modular Multilevel Converter (MMC) is considered as an attractive choice to be used in high-voltage applications. Despite significant advantages of Finite Control Set Model Predictive Control (FCS-MPC) scheme, notable computational load is the main obstacle for implementing FCS-MPC technique for MMCs with high number of submodules. In order to solve this issue, an indirect two-stage FCS-MPC strategy with highly-reduced computations, is proposed in this paper. As an indirect FCSMPC approach, only the output and circulating currents are considered in the cost function, and an external voltage balancing algorithm should be employed to determine the submodule operating states, based on the MMC arm voltage levels. In the first stage of proposed scheme, the initial arm voltage references are calculated, according to the discrete MMC model, and references of output and circulating currents. In the second stage, the admissible arm voltage levels for cost function optimization, are restricted to vicinity of the references in first stage. In this way, the number of times that the cost function is evaluated, would become independent of the number of submodules. Finally, the performance of proposed scheme is validated by simulation results.