A Hybrid MPC-PI Control Strategy for Grid-Connected Modular Multilevel Converters Using LS-PWM and Y-Matrix State Selection

In the grid-connected Modular Multilevel Converter (GC-MMC) system, it is challenging to simultaneously achieve three goals, such as suppressing circulating current, balancing capacitor voltage, and regulating grid current, due to the complexity of the control system. The integration of pulse width modulation (PWM) techniques with model predictive control (MPC) in GC-MMC system is a very challenging task, which requires complicated duty cycle calculation process. It also necessitates the use of additional sorting algorithms to find the optimal switching states. Considering these issues, this paper proposes a hybrid algorithm based on MPC and PI (HMPC-PI) control with a simple level shifted PWM (LS-PWM) and Y-matrix based state selection approach. The idea behind the proposed method is to employ the LS-PWM concept to generate the control logic for grouping the switching states in Y-matrix row selector in accordance with the steps of the output voltage produced by MMC. The proposed controller generates optimal PWM signals by merely selecting switching state from each group in the Y-matrix. With the help of HMPC-PI algorithm, it is possible to formulate the optimal control problem to meet multiple control objectives. The effectiveness of the HMPC-PI algorithm is validated using extensive simulations.