Review on computational methods for Lyapunov functions (Q258388)

In 1892 Lyapunov published his doctoral dissertation in which he introduced a sufficient condition for the stability of a nonlinear system, namely, the existence of a positive definite function decreasing along the solution trajectories. These functions are now known as Lyapunov functions. This review paper is divided in two parts, the first discussing different Lyapunov and Lyapunov-type functions for various types of systems, and the second discussing numerical methods for computing Lyapunov functions. Besides finite-dimensional systems of ordinary differential equations, Lyapunov functions have been developed for discrete-time systems (maps), PDEs, non-autonomous systems of ODEs, switched systems, delay differential equations, control systems and random dynamical systems, among others. Generalisations of Lyapunov functions include non-smooth Lyapunov functions, and vector, matrix and multiple Lyapunov functions. Numerical methods used to compute Lyapunov functions include collocation, linear programming, linear matrix inequalities, algebraic methods and graph theoretic methods. In this review the authors bring together these different methods and describe the state of the art for the vast variety of methods to compute Lyapunov functions for various kinds of systems.