Radial basis function neural network-based adaptive control of uncertain nonlinear systems

We aim to design in the present paper an adaptive output feedback control scheme to address the tracking problem of an uncertain system having full relative degree in the presence of neglected dynamics and modelling errors. Then, the obtained controller is augmented by an online radial basis function neural network (RBF NN) that is used to adaptively compensate for the nonlinearity existing in the uncertain systems. A linear observer is introduced to generate an error signal for the adaptive laws. Ultimate boundedness is proven through Lyapunov's direct method. The forcefulness of the theoretical results is demonstrated through computer simulations of a nonlinear second-order system.