基于事件触发控制的中立型神经网络簇同步  

Cluster Synchronization of Neutral Neural Networks Based on Event-triggered Control

作　　者：邓伟伟 高燕[1] 王春媛 DENG Weiwei;GAO Yan;WANG Chunyuan(School of Electronic and Electrical Engineering,Shanghai University of Engineering Science,Shanghai 201600,China)

机构地区：[1]上海工程技术大学电子电气工程学院,上海201600

出　　处：《控制工程》2024年第2期253-261,共9页Control Engineering of China

基　　金：国家自然科学基金资助项目(61801286)。

摘　　要：研究了事件触发下带有Lévy噪声的中立型耦合神经网络的簇同步问题。首先,在混合时滞和不连续噪声的条件下,提出了一种事件触发机制与牵制控制方法结合的控制策略,应用于神经网络的簇同步中,以提高簇同步效率,缓解网络负担。然后,通过构建新的Lyapunov泛函,以及使用线性矩阵不等式(linear matrix inequalities,LMI)分析技术和广义Dynkin公式,得到误差系统的簇同步准则,进一步保证了耦合神经网络的簇同步。最后,通过一个数值仿真,验证了所得结论的可行性。The cluster synchronization problem for neutral coupled neural networks with Lévy noise via an event-triggered mechanism is investigated.Firstly,under the condition of mixed delays and discontinuous noise,a control strategy combined event-triggered mechanism and pinning control method is proposed,which is applied to the cluster synchronization of neural network to improve the cluster synchronization efficiency and alleviate the network burden.Then,by constructing a new Lyapunov functional and using LMI analysis technology and generalized Dynkin formula,the cluster synchronization criterion of the error system is obtained,which can further ensure the cluster synchronization of the coupled neural network.Finally,the feasibility of the conclusion is verified by a numerical simulation.

关 键 词：中立型神经网络 Lévy噪声 牵制控制 事件触发 簇同步 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]