Optimal Model-Free Tracking Control Based on Direct Adaptive Fuzzy Controller for a Mini-Drone

This paper presents a direct robust adaptive fuzzy controller designed for a mini drone. The controller is developed to address the challenge of controlling the six-coupled nonlinear subsystems of the mini-drone with unknown dynamics for position and attitude control. The proposed control approach combines a fuzzy adaptive controller and the Bat Algorithm (BA) based optimization technique to elaborate decentralized control structure. The control inputs of the drone are generated through a Mamdani-type fuzzy adaptive controller, based exclusively on the system's tracking errors. This design choice effectively circumvents the singularity issues commonly encountered in indirect adaptive control schemes, while avoiding the need for robust state observers. To address external disturbances and model uncertainties, a robust compensatory term is incorporated into the fuzzy controller. The Bat Algorithm (BA) is employed to optimize the adaptation parameters of the controller to achieve the best performance in trajectory tracking. This approach resolves the trial-and-error issue in parameter selection which gives a systematic and optimal tuning of the control parameters. The stability of the closed-loop system is rigorously established via Lyapunov theory. The effectiveness of the proposed controller is demonstrated through various simulation cases on a virtual platform of Parrot Mini-Drone, as well as experimental validation for hovering. The results show that the proposed controller can handle unknown dynamics, uncertainties, and external disturbances while achieving accurate position and attitude control.