Extended Dissipativity and Non-Fragile Synchronization for Recurrent Neural Networks With Multiple Time-Varying Delays via Sampled-Data Control

R Anbuvithya,R Vadivel,Nallappan Gunasekaran,Porpattama Hammachukiattikul,S Dheepika Sri

doi:10.1109/access.2021.3060044

Abstract

This paper deals with the extended dissipativity and non-fragile synchronization of delayed recurrent neural networks (RNNs) with multiple time-varying delays and sampled-data control. A suitable Lyapunov-Krasovskii Functional (LKF) is built up to prove the quadratically stable and extended dissipativity condition of delayed RNNs using Jensen inequality and limited Bessel-Legendre inequality approaches. A non-fragile sampled-data approach is applied to investigate the problem of neural networks with multiple time-varying delays, which ensures that the master system synchronizes with the slave system and is designed with respect to the solutions of Linear Matrix Inequalities (LMIs). The effectiveness of the suggested approach is established by providing suitable simulations using MATLAB LMI control toolbox. Finally, numerical examples and comparative results are provided to illustrate the adequacy of the planned control scheme.

Highlights

Neural Networks (NNs) provide an interesting pattern for a wider extent of complex systems over the past few decades
In the following Theorems, we design the non-fragile sampled-data control and some sufficient conditions are given to guarantee that the error system (6) and (7) is synchronizes and extended dissipative in the form of Linear Matrix Inequalities (LMIs)
In this paper, the extended dissipativity of Recurrent neural networks (RNNs) with multiple time-varying delays has been studied via non-fragile sampled-data control

Summary

INTRODUCTION

Neural Networks (NNs) provide an interesting pattern for a wider extent of complex systems over the past few decades. R. Anbuvithya et al.: Extended Dissipativity and Non-Fragile Synchronization for RNNs. uncertainty for stochastic discrete-time systems has been proposed. The idea of reliable asynchronous sampled-data filtering of T-S fuzzy uncertain delayed neural networks with stochastic switched topologies has been considered in [27]. Different from most of the published works, the problem of non-fragile synchronization for extended dissipativity and sampled-data RNNs with multiple time-varying delays are studied here. The main contributions and novelty of this paper are summarised as follows: (1)The quadratic stability and extended dissipativity of RNNs with multiple time-varying delays are obtained. (3) The desired non-fragile sampled-data controller for the considered system can be obtained with respect to a new set of LMIs utilizing MATLAB toolbox.

PRELIMINARIES

MAIN RESULTS

SIMULATION RESULTS

CONCLUSION