Adaptive RBF Neural Network with sliding mode controller for a nonlinear flexible wing aircraft

Abstract

Our research will concern the study of an adaptive neural network sliding mode controller using radial basic function to reduce the aeroelastic instabilities of a nonlinear aeroelastic system. The strategy of the designed controller is to provide the ability to damp rapidly the high amplitude oscillation that occurred in the system and to enhance the flight conditions beyond the open-loop critical flutter speed. Also, the proposed controller can be used to estimate the model and its nonlinear dynamics. The two-degree-of-freedom nonlinear aeroelastic system describes the pitch and plunge motion of the aircraft wing section equipped with leading- and trailing edge control surfaces. Furthermore, the selected model considers the quasi-steady aerodynamic model and the structural stiffness nonlinearities. The objective of the simulations is to show the behavior of the system in open and closed-loop and to demonstrate the ability of the controller to reduce the oscillation amplitude in the subcritical flight speed range and to derive the state trajectories to the origin even in t he presence of the disturbances and the uncertainties.