Adaptive exponential synchronization of chaotic recurrent neural networks with stochastic perturbation

Abstract

This paper presents an adaptive exponential synchronization scheme for chaotic recurrent neural networks with stochastic perturbation. An adaptive synchronization controller is developed based on linear matrix inequality (LMI) and the controller can guarantee that the error of synchronization is exponentially ultimately bounded stable in the mean square. Moreover, we can make the bound of error as small as possible by appropriate selections of the controller parameters. Finally, an example is given to illustrate the validity of the proposed design.