Improved Even-Handed Predictive Current Control for Modular Multilevel Converters

Yafei Yin,Zhenbin Zhang,Minghao Han,Yuanxiang Sun,Zhen Li
DOI: https://doi.org/10.1109/precede51386.2021.9680923
2021-01-01
Abstract:Model predictive control (MPC) becomes a very promising alternative for modular multilevel converters (MMCs). Conventional MPC solutions design sequential control structures for grid filter current and MMC circulating current to reduce calculation burden. However, how to choose the priority becomes an open question in such structures. Besides, sequential structures are incapable to include a custom-designed cost function, and hence hard to meet user demands. To tackle the problems, this work proposes an improved even-handed predictive current control. The proposed method selects two candidate voltage levels for each phase (equivalents to eight candidate voltage vectors for three phases), with a minimum cross-error criterion. Afterward, the optimal voltage vector will be selected by minimizing a custom-designed cost function with a three-phase MMC prediction model. Since even-handed selection results are calculated directly according to references, the proposed method has a low calculation burden. Simulation results, at different test scenarios, confirm the effectiveness of the proposal.
What problem does this paper attempt to address?