Continuous Control Set Model Predictive Control of a Hybrid Modular Multilevel Converter for Wind Energy Applications

This article discusses a continuous control set model predictive control system of a hybrid modular multilevel converter, controlling a direct drive permanent magnet synchronous generator, for variable-speed wind energy applications. The hybrid converter enables operation at a reduced dc-link voltage without the need for circulating currents, even during low-frequency operation in the ac port. To further reduce cluster currents, the capacitor voltages are allowed to oscillate inside predefined voltage limits. The control system is based on a single-step predictive control algorithm, whose outputs are the dc-link and common-mode voltages required to balance the converter energy. Furthermore, it is demonstrated in this work that the reference for the common-mode voltage, obtained using an algorithm based on the Moore-Penrose pseudoinverse matrix, is a third harmonic waveform, which can be used to balance the capacitor voltages, as well as to increase the modulation index. The proposed control system is validated using a hardware-in-the-loop platform and an experimental system consisting of a 5-kW back-to-back hybrid modular multilevel converter, controlled using dSPACE MicroLabBox platforms.