Abstract <p>The present paper suggests a sophisticated control strategy for a boost converter, which is a naturally nonminimum phase system based on a fractional-order adaptive-robust controller (FOARC) tuned by the use of the gray wolf optimization (GWO) algorithm and Q-learning. Non-minimum phase converters, such as the boost converter, are incredibly challenging regarding stability, transient response, and performance through their intrinsic time-delay characteristics and inverse dynamics. The complexities involved make traditional control methods fail to obtain quick response and precise tracking. The flexibility of fractional-order elements is used in the proposed FOARC to advance the converter’s dynamic performance through better control of system response compared with traditional integer-order controllers. The GWO and Q-learning algorithms are employed to optimize the controller’s parameters. The GWO algorithm is used in the global parameter space search to ensure a well-distributed optimization. At the same time, the Q-learning fine-tunes the controller’s performance by adapting it to change, guaranteeing a constant performance improvement. Simulation outcomes reveal that the integration of GWO and Q-learning in optimizing the FOARC is superior to the conventional methods in transient response, steady-state error minimization, and system robustness. The proposed control strategy offers a possible solution for high-performance power conversion systems, particularly when demanding precision and efficiency.</p>