An Enhanced Model Predictive Capacitor Voltage Control of Hybrid Modular Multilevel Converters Under Over-Modulation Circumstances

Modular multilevel converters (MMCs) require well-balanced submodule (SM) capacitor voltages at the reference value for satisfactory steady-state and dynamic operations. Achieving this goal in hybrid MMCs, where the modulation index exceeds one, becomes intricate owing to the differing participation of half-bridge SMs (HB-SMs) and full-bridge SMs (FB-SMs) in arm voltage generation. This paper proposes an advanced voltage balancing method based on enhanced phase-shift modulation and model predictive control, featuring a single comprehensive cost function that utilizes a novel form of redundancy to manage the insertion of HB- and FB-SMs in a wide range of modulation indices. The designed cost function addresses four key control objectives: inter-cell balancing by tracking the DC voltage reference, minimizing the energy deviation between the HB- and FB-SMs, inter-leg balancing to eliminate energy variations between the upper and lower arms, and rejecting voltage errors arising from uncertainties in the SM capacitance using a Kalman filter (KF)-based observer. This method enables comprehensive capacitor voltage balancing, which is suitable for various modulation indices and power factors. The validity of the proposed method is substantiated through simulations using MATLAB/SIMULINK software and experiments conducted on a scaled-down hardware prototype.