Yang Xinsong, Cao Jinde, Qiu Jianlong
Department of Mathematics, Chongqing Normal University, Chongqing, 401331, China.
Department of Mathematics, Southeast University, Nanjing 210096, China; Department of Mathematics, Faculty of Science, King Abdulaziz University, Jeddah 21589, Saudi Arabia.
Neural Netw. 2015 May;65:80-91. doi: 10.1016/j.neunet.2015.01.008. Epub 2015 Feb 4.
This paper concerns the pth moment synchronization in an array of generally coupled memristor-based neural networks with time-varying discrete delays, unbounded distributed delays, as well as stochastic perturbations. Hybrid controllers are designed to cope with the uncertainties caused by the state-dependent parameters: (a) state feedback controllers combined with delayed impulsive controller; (b) adaptive controller combined with delayed impulsive controller. Based on an impulsive differential inequality, the properties of random variables, the framework of Filippov solution, and Lyapunov functional method, sufficient conditions are derived to guarantee that the considered coupled memristor-based neural networks can be pth moment globally exponentially synchronized onto an isolated node under both of the two classes of hybrid impulsive controllers. Finally, numerical simulations are given to show the effectiveness of the theoretical results.
本文研究了一类具有时变离散延迟、无界分布延迟以及随机扰动的基于忆阻器的广义耦合神经网络阵列中的第(p)阶矩同步问题。设计了混合控制器来应对由状态依赖参数引起的不确定性:(a) 状态反馈控制器与延迟脉冲控制器相结合;(b) 自适应控制器与延迟脉冲控制器相结合。基于脉冲微分不等式、随机变量的性质、菲利波夫解的框架以及李雅普诺夫泛函方法,推导出了充分条件,以确保在这两类混合脉冲控制器下,所考虑的基于忆阻器的耦合神经网络能够在第(p)阶矩意义下全局指数同步到一个孤立节点。最后,给出了数值模拟以验证理论结果的有效性。