Synchronization of Neural Networks with Mixed Time-Varying Delays Based on Parameter Identification and via Output Coupling

Abstract

This paper aims to investigate the global robust synchronization problem for two coupled neural networks with both discrete and distributed time-varying delays via output coupling. A general and novel time-varying delayed feedback scheme is introduced to model a more realistic controller. By employing the Lyapunov stability theory, several new and less restrictive criterions are obtained to guarantee that the two coupled chaotic neural networks can achieve synchronization. In addition, each adapted parameter in the connection weights can be identified through the theoretical results. Numerical simulations are given to validate the usefulness of the proposed global synchronization conditions.