The negative impedance behavior of constant power loads (CPLs) often destabilizes DC microgrids by reducing the damping effect. Additionally, non-conventional disturbances can negatively affect DC–DC buck–boost converters in DC microgrids, causing severe voltage oscillations. To overcome these challenges, this paper proposes a generalized predictive controller (GPC) with an adaptive fuzzy-based prediction horizon for a DC–DC buck–boost converter connected to a constant power load under disturbances. The adaptive prediction horizon improves control performance when facing non-conventional disturbances. Furthermore, since predictive control depends on the nominal model of the system and real systems are subject to disturbances, a fixed-time disturbance observer (FTDO) is integrated into the controller to identify and counteract these disturbances. To demonstrate the effectiveness of the proposed controller, a comprehensive comparison is conducted among the PI controller, the fixed- horizon predictive controller, the self-tuning adaptive predictive controller, and the fuzzy-based adaptive predictive controller. The superiority of the fuzzy-based approach is evaluated in terms of dynamic performance. Simulation results in MATLAB/Simulink confirm the performance improvement and effectiveness of the proposed control strategy. The practical effectiveness of the proposed control algorithm is further validated through processor-in-the-loop (PIL) testing using an STM32 Nucleo-G431RB board.
