<p>This article aims to present a learning-based robust optimal sequential control framework for DC–DC buck converters operating under uncertain conditions, specifically with respect to variations in resistive loads. The proposed control strategy employs an output-feedback design, utilizing only the capacitor voltage measurement for feedback purposes. To address system uncertainties, a fuzzy system is implemented to approximate unknown term, and a new robust fuzzy observer is developed for estimation of the inductor current. This observer integrates a disturbance descriptor to enhance robustness against uncertainties. Furthermore, a command filtering unit is incorporated to streamline the controller design process, and adaptation rules are formulated to update the fuzzy logic weights. Parameter optimization is also achieved through the utilization of the particle swarm optimizer of fractional-order. Stability analysis, conducted via Lyapunov synthesis, verifies the uniformly ultimate boundedness of all signals within the entire system. Comprehensive simulations are presented to evaluate the system’s response under varying input voltage and load resistance. The efficacy of the proposed idea is confirmed, with experimental data substantiating the theoretical findings.</p>