${H}_{∞ }$ Bipartite Synchronization of Double-Layer Markov Switched Cooperation-Competition Neural Networks: A Distributed Dynamic Event-Triggered Mechanism
Jing Wang,Mengping Xing,Jinde Cao,Ju H. Park,Hao Shen
DOI: https://doi.org/10.1109/tnnls.2021.3093700
IF: 14.255
2021-01-01
IEEE Transactions on Neural Networks and Learning Systems
Abstract:In this article, the H<sub>∞</sub> bipartite synchronization issue is studied for a class of discrete-time coupled switched neural networks with antagonistic interactions via a distributed dynamic event-triggered control scheme. Essentially different from most current literature, the topology switching of the investigated signed graph is governed by a double-layer switching signal, which integrates a flexible deterministic switching regularity, the persistent dwell-time switching, into a Markov chain to represent the variation of transition probability. Considering the coexistence of cooperative and antagonistic interactions among nodes, the bipartite synchronization of which the dynamics of nodes converge to values with the same modulus but the opposite signs is explored. A distributed control strategy based on the dynamic event-triggered mechanism is utilized to achieve this goal. Under this circumstance, the information update of the controller presents an aperiodic manner, and the frequency of data transmission can be reduced extensively. Thereafter, by constructing a novel Lyapunov function depending on both the switching signal and the internal dynamic nonnegative variable of the triggering mechanism, the exponential stability of bipartite synchronization error systems in the mean-square sense is analyzed. Finally, two simulation examples are provided to illustrate the effectiveness of the derived results.
computer science, artificial intelligence, theory & methods,engineering, electrical & electronic, hardware & architecture