A Novel Plug-and-Play Cooperative Disturbance Compensator for Heterogeneous Uncertain Linear Multi-Agent Systems

Abstract

Cooperative output regulation (COR) for multi-agent systems (MAS) has garnered significant attention due to its broad applications. This letter offers a fresh perspective on the COR problem for a class of heterogeneous, uncertain, linear SISO MAS facing two major challenges simultaneously: (1) the agents are highly uncertain and heterogeneous, and (2) communication is restricted to a directed spanning tree with only local information exchanged among agents. We propose a novel plug-and-play cooperative feedforward disturbance compensator that requires minimal prior knowledge of follower agents’ dynamics. In contrast to traditional methods, our compensator is fully distributed, adaptive, and highly robust to agent heterogeneity. It eliminates the need for system identification and handles large uncertainties without relying on typical assumptions such as minimum phase, identical dimensionality, or uniform relative degree across agents. Additionally, the compensator is designed for scalability, offering plug-and-play functionality that allows seamless addition or removal of agents without requiring controller redesign, provided the network maintains a spanning tree. Theoretical analysis and simulations demonstrate the compensator’s effectiveness in solving the COR problem across various scenarios.