Predefined-time synchronization of competitive neural networks with deterministic disturbances and stochastic noises

Abstract

This paper presents the drive-response synchronization problem for competitive neural networks in predefined time. The response system is considered in the presence of deterministic disturbances satisfying Lipschitz conditions and in the presence of both stochastic white noises and deterministic disturbances satisfying Lipschitz conditions. The effect of deterministic disturbances and stochastic noises is suppressed by designing a linear time-varying continuous control input driving the synchronization errors at the origin for an a priori predefined time, independently of initial conditions, deterministic disturbances, and stochastic noises. Finally, numerical simulations are conducted to demonstrate validity of the obtained theoretical results. The conducted comparisons with other predefined-time convergent synchronization algorithms reveal better performance of the proposed synchronization technique.