Pareto-based Optimised Weighting Factor for Model Predictive Control for Flying Capacitor Converter

The multiple control objective, balancing of flying capacitor (FC) voltages and the output current sets the challenge to controller design for the FC converter. The finite-control-set model predictive control (MPC) has emerged as a popular control method for the FC converter. However, the original cost function design often faces the issue of selecting suitable weighting factors. To address this challenge, this paper proposes an offline and an online strategy that utilise the principle of Pareto optimality and a knee-based Pareto pruning method. The offline approach uses the non-dominated sorting genetic algorithm (NSGA-II) to explore the optimal weighting factors. In contrast, the online strategy employs the knee-based decision-making method that avoids weighting factors. Simulation results are provided to validate the feasibility of both approaches.