Adaptive H∞ control for a class of non-linear systems using neural networks

Abstract

An adaptive H∞ controller is designed for a class of non-linear dynamical systems by using neural networks (NNs) to approximate the unknown system function. It is shown that under bounded initial conditions, the H∞ performance from the external disturbance to the controlled output is achieved with a prescribed attenuation level. A systematic design procedure is then developed for the synthesis of the adaptive H∞ controller by solving two linear matrix inequalities (LMIs). Finally, a numerical example is included to demonstrate the effectiveness of the proposed controller.