Capacitor voltage estimation in modular multilevel converters using a Kalman filter algorithm

This paper proposes an adaptive observer based on Kalman filter (KF) to estimate the capacitor voltages of Modular Multilevel Converters (MMC). The proposed method offers a considerable reduction in sensors count, hence the system cost and also increases the converter reliability. The paper also investigates several approaches linked to MMC; voltage balancing control issue, the effect of circulating current on the arm capacitors voltages in addition to sensors cost issues. The effectiveness of the proposed method is validated through simulation studies of a single-phase five-level MMC using MATLAB/SIMULINK.