Non-weighted H-infinity state estimation for discrete-time switched neural networks with persistent dwell time switching regularities based on Finsler's lemma

摘要

In this study, the state estimation and Ho. control problem for discrete-time switched neural networks with mode-dependent time-varying delays has been studied with persistent dwell time (PDT) switching regularities. The phenomenon of PDT, existing for the designed estimator of underlying switched neural networks are characterized by introducing a Bernoulli distributed white sequence. The main aim of the addressed problem is to design mode dependent state estimators such that the dynamics of the estimation error is exponentially stable with an expected decay rate and satisfies the prescribed H-infinity performance constraint. Sufficient conditions are established for the occurence of the desired filter to ensure the mean-square exponential stability of the augmented system by using the generalized Finsler's lemma and then the full-order filter parameters are presented in terms of solutions to a set of linear matrix inequality (LMI) conditions. Finally, simulation results are given to explain the usefulness of the proposed design procedure. (C) 2017 Elsevier B.V. All rights reserved.