Data-Driven Cooperative Output Regulation Via Output Feedback Control

The cooperative output regulation problem for multi-agent systems (MASs) serves as a foundational framework for various cooperative control tasks. While extensively studied in model-based settings, its data-driven counterpart under output feedback remains largely unexplored. This paper addresses this gap by developing a distributed data-driven output feedback control scheme grounded in the internal model principle. We first derive a representation of the original MAS using historical input-output data. By embedding an internal model into this representation, we introduce an augmented system that reformulates the cooperative output regulation problem into a stabilization problem. A key contribution is the development of agent-wise semidefinite programs (SDPs) that compute stabilizing controllers directly from data, bypassing output regulator equations (OREs) and system identification. Theoretical guarantees for asymptotic stability and exact tracking are rigorously established. Numerical experiments validate the method's efficacy in both standard and challenging scenarios.