Dual-Model Artificial Neural Network based control for a 4-level Flying Capacitor Converter

This paper presents the design, training, and implementation of a Feed-Forward Neural Network based on a dual control strategy to manage a four-level Flying Capacitor Converter. The dual control strategy comprises Proportional-Integral (PI) control and Finite Control Set Model Predictive Control (FCS-MPC). The proposed approach enables switching between those strategies based on the system’s operational state, utilizing a hysteresis band, where PI control is activated during steady-state, while FCS-MPC is employed during transient, improving dynamic performance. The primary objective is to harness the benefits of both control strategies by implementing an ANN. This will provide an optimal solution in terms of performance and computational cost while ensuring smooth transitions and minimizing disturbances during state changes.