State Estimation for Markovian Coupled Neural Networks with Multiple Time Delays Via Event-Triggered Mechanism

This paper focuses on the state estimation problem for a type of coupled neural networks with multiple time delays and markovian jumping communication topologies. To avoid unnecessary resources consuming, a novel state estimator is designed based on event-triggered mechanism, in which the control input of each node is only updated when the measurement output error exceeds a predefined threshold. The event-triggering time sequence is a subset of the switching time sequence, which can naturally excludes the Zeno-behavior. By utilizing an appropriate Lyapunov-Krasovskii functional, as well as the weak infinitesimal operator of Markov process and some algebraic inequalities, an easy-to-check sufficient criterion is derived to ensure the exponential ultimate boundedness of the estimation error. Finally, a simulation example is presented to illustrate the applications and effectiveness of the theoretical results.

Automated summary

This paper introduces a novel state estimator based on event-triggered mechanism to address the state estimation problem of coupled neural networks, which effectively reduces resource consumption and eliminates unnecessary Zeno behavior. By utilizing an appropriate Lyapunov-Krasovskii functional and the weak infinitesimal operator of Markov process, a sufficient criterion is derived to ensure the exponential ultimate boundedness of the estimation error.