Robust Synchronization for Discrete-Time Coupled Markovian Jumping Neural Networks With Mixed Time-Delays

Abstract

This paper concerns the robust synchronization problems for discrete-time coupled neural networks with discrete time delay and distributed time delays. Inner parameters in individual neural network are subject to be uncertain and both coupled matrixes and weight matrixes are supposed to switch from one mode to another because of the markovian jumping chain. Mixed time delays contain discrete and distributed time delays and the mixed time delays not only exist in the individual neural cell, but also exist in the coupled cells. By using the novel Lyapunov-Krasovskii functional method and Kronecker product as tools, mean square stability conditions are provided in terms of linear matrix inequalities. In numerical simulations, two examples (with and without unknown parameters) are given and simulation results show the robustness and effectiveness of our methods.

Keywords

• Discrete-time coupled neural networks
• Markovian jumping chain
• mixed time delays
• linear matrix inequality
• Lyapunov-Krasovskii functional method