A hybrid method for the analysis of non-uniformly sampled systems

Abstract

In this chapter we propose a method for the analysis of sampled-data systems with sampling jitter. We consider that the sampling interval is unknown and time-varying and we provide a method for estimating the Lyapunov exponent. The proposed method is hybrid, in the sense that it combines continuous-time models (based on time delay systems) with polytopic embedding methods, specific to discrete-time approaches. The approach exploits the fact that the command is a piecewise constant signal and leads to less conservative stability conditions with respect to the existing literature. Using geometrical arguments, a lower bound of the Lyapunov exponent can be expressed as a generalized eigenvalue problem. © 2012 Springer-Verlag GmbH Berlin Heidelberg.