Development of Hyperplane-based Adaptive T-S Fuzzy Controller for Micro Aerial Robots

Abstract

In recent times, numerous applications of autonomous systems are witnessed vividly. Efficient control of autonomous systems like micro aerial robots (MARs) is challenging since their dynamics is highly nonlinear and associated with uncertainties. Therefore, an increasing interest is noticed in developing adaptive and computationally effective intelligent controllers. In this work, a hyperplane-based adaptive Takagi-Sugeno (TS) fuzzy controller namely hyperplane-based adaptive fuzzy (HPAF) controller is developed. Unlike the existing adaptive fuzzy controller, HPAF is characterized by fewer system parameters since it has no antecedent parameters. Such feature yields a fast response to follow the desired control commands in a challenging environment. To observe a sharp convergence of tracking error to zero, the consequent parameters of the HPAF are tuned through adaptation laws derived from a radial basis function neural network (RBFNN). Our HPAF controller's closed-loop stability has also been proved using Lyapunov theorem. Finally, the proposed controller's performance has been evaluated by employing it to control the altitude of a bioinspired flapping wing MAR and compared with a proportional integral derivative (PID) and static TS-Fuzzy controller, where better tracking performances are perceived than the benchmark controllers.