Event-Triggered Bipartite Synchronization of Delayed Inertial Memristive Neural Networks With Unknown Disturbances
Xiaoyang Liu, Haibin He, Jinde Cao
Abstract
This paper focuses on the quasi-bipartite synchronization problems of delayed inertial memristive neural networks (DIMNNs) with signed graphs and unknown external disturbances. Firstly, an event-triggered hybrid impulsive mechanism is proposed to save the communication resources. In light of average impulsive interval theory and comparison principle, the range of impulsive effects is discussed so that neither positive nor negative effects disrupt the synchronization of the networks. Secondly, with the help of neural network (NN) approximation theory, a neuro-adaptive term is developed to resist the effects of unknown disturbances, and several conditions are given for quasi-bipartite synchronization. Furthermore, the lower bound of the event-triggered intervals demonstrates that Zeno behaviors can be avoided. Lastly, the validity of the designed protocol is substantiated by a numerical example.