Asynchronous Sampled-Data State Estimation for a Class of Nonlinear Complex Networks: A Matrix-Exponential-Gain-Based Approach

Abstract

This paper is concerned with the asynchronous sampled-data state estimation problems for a class of continuous-time nonlinear complex networks. A novel asynchronous sampled-data estimator is constructed with matrix exponential gains to estimate the states of network nodes, which allows each node to independently sample and transmit the measured signals at its own designated time instants. It is demonstrated that the utilization of matrix exponential gains is capable of enlarging the maximum-allowable bound of the sampling intervals. Moreover, a modified Halanay-type inequality is derived to facilitate the analysis of estimation errors. Accordingly, by leveraging the Lyapunov stability theory, some sufficient conditions are obtained to guarantee the global exponential stability of the estimation error dynamics. In addition, the maximum-allowable bound of the sampling intervals is explicitly characterized by resorting to an algebraic inequality, and a convex optimization method is adopted with the aim of maximizing such an allowable bound. Finally, some numerical simulations are conducted to validate the feasibility and usefulness of the established theoretical results.