A Novel Clustering Approach Based on Group Quasi-Consensus of Unstable Dynamic Linear High-Order Multi-Agent Systems

TL;DR

This paper proposes a new clustering method based on group consensus in dynamic high-order multi-agent systems, linking graph topology with agent behavior to identify clusters.

Contribution

It introduces a novel clustering approach grounded in group consensus theory for high-order multi-agent systems, with a new necessary and sufficient condition for consensus.

Findings

The approach effectively reveals cluster structures in simulated examples.

Theoretical conditions for group consensus are established.

Numerical instances demonstrate the method's applicability.

Abstract

This paper introduces a novel approach of clustering, which is based on group consensus of dynamic linear high-order multi-agent systems. The graph topology is associated with a selected multi-agent system, with each agent corresponding to one vertex. In order to reveal the cluster structure, the agents belonging to a similar cluster are expected to aggregate together. As theoretical foundation, a necessary and sufficient condition is given to check the group consensus. Two numerical instances are shown to illustrate the process of approach.