Distributed Scalable Control for Large-Scale Networked Systems Using Chordal Decomposition

Abstract

This letter proposes a distributed scalable control scheme for large-scale networked systems using chordal decomposition. To overcome the limitations caused by coupled subsystem structures, we develop a modeling framework for large-scale networked systems with arbitrary topological connections over directed graphs. Subsequently, leveraging the chordal decomposition theorem, we derive a sufficient condition for distributed scalable stabilization of such systems. Unlike traditional centralized or distributed control methods, the proposed approach inherently supports plug-and-play operations. Specifically, it eliminates the need to redesign all controller gains when the system undergoes partial modifications. Furthermore, the method improves computational efficiency while maintaining data confidentiality. Finally, the effectiveness of the proposed approach is validated through a coupled microgrid test system.