Design of an event-triggered extended dissipative state estimator for neural networks with multiple time-varying delays

Abstract

This paper examines the issue of designing an extended dissipative state estimator for a class of neural networks with multiple time-varying delays. The novelty of this problem lies in assuming distinct time-varying delays for each node, demonstrating its generalizability and complexity. An event-triggered state estimator with a known output measurement is proposed to facilitate these targeted network responses by saving limited communication resources. Consequently, sufficient conditions for an extended dissipative estimator have been achieved by constructing an augmented Lyapunov–Krasovskii functional (LKF) and finding its derivative. A generalized free-weighting matrix inequality (GFWMI) is utilized to achieve a tighter upper bound of the derivative, leading to a less conservative result in linear matrix inequalities (LMIs). Ultimately, a numerical example is shown to verify the advantages and efficacy of the main findings.