Event-Based Impulsive Control for Heterogeneous Neural Networks with Communication Delays

The quasi-synchronization for a class of general heterogeneous neural networks is explored by event-based impulsive control strategy. Compared with the traditional average impulsive interval (AII) method, instead, an event-triggered mechanism (ETM) is employed to determine the impulsive instants, in which case the subjectivity of selecting the controlling sequence can be eliminated. In addition, considering the fact that communication delay is inevitable between the allocation and execution of instructions in practice, we further nominate an ETM centered on communication delays and aperiodic sampling, which is more accessible and affordable, yet can straightforwardly avoid Zeno behavior. Hence, on the basis of the novel event-triggered impulsive control strategy, quasi-synchronization of heterogeneous neural network model is investigated and some general conditions are also achieved. Finally, two numerical simulations are afforded to validate the efficacy of theoretical results.

Automated summary

This paper explores the quasi-synchronization of a general class of heterogeneous neural networks using an event-based impulsive control strategy. Instead of the traditional average impulsive interval (AII) method, an event-triggered mechanism (ETM) is employed to determine the impulsive instants, eliminating the subjectivity in selecting the controlling sequence. Furthermore, considering the inevitable communication delay between instruction allocation and execution, an ETM centered on communication delays and aperiodic sampling is proposed to avoid Zeno behavior.