Convergence of Opinion Dynamics With Heterogeneous Asymmetric Saturation Levels and Time-Varying Delays

Yao Zou, Wei Wang, Kewei Xia, Zongyu Zuo*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Saturation and delay constraints, which are inevitable in reality, cause severe uncertainty to the behaviors of network dynamics. Accordingly, this paper studies the combined saturation and delay effect in the convergence of opinion dynamics over a switching cooperative-antagonistic network with weak connectivity. In particular, the considered saturations are in a generic form that the saturation level of each agent is asymmetric and the saturation levels of different agents are non-identical. And the delays of different channels are considered heterogeneous and time-varying. It is first demonstrated that the typical bipartite consensus result associated with the opinion dynamics is of benign tolerance to the saturation and delay constraints given a uniformly jointly strongly connected and structurally balanced topology condition. Next, by relaxing the uniformly jointly strongly connected condition just to roots, it is shown that the bipartite consensus is still ensured by the structural balance given a uniform joint reachability condition for non-roots. Besides, if the structural balance of the non-roots is absent, their opinions are shown to be ultimately bounded by the polarized values of the roots. Finally, two examples are simulated to validate the built theoretical results.

Original languageEnglish
Pages (from-to)3189-3198
Number of pages10
JournalIEEE Transactions on Network Science and Engineering
Volume11
Issue number3
DOIs
Publication statusPublished - 1 May 2024

Keywords

  • Opinion dynamics
  • Structural balance
  • bipartite. consensus
  • cooperative-antagonistic network
  • delays
  • saturations

Fingerprint

Dive into the research topics of 'Convergence of Opinion Dynamics With Heterogeneous Asymmetric Saturation Levels and Time-Varying Delays'. Together they form a unique fingerprint.

Cite this