Generation of Lyapunov Functions by Neural Networks
Navid Noroozi, Paknoosh Karimaghaee, Fatemeh Safaei, Hamed Javadi · 2008
Abstract—Lyapunov function is generally obtained based on trail and error. Systems with high complex nonlinearities and more dimensions are much more difficult to be dealt with and in many cases it leads to dull one. This paper proposes two straightforward methods for determining or approximating a Lyapunov function based on approximation theory and the abilities of artificial neural networks. The potential of the proposed methods are demonstrated by simulation examples. Index Terms—Approximation theory, Lyapunov function, nonlinear system, neural network.