Robust Data-Driven Control of Heterogeneous Multi-Agent Systems and Its Application in Autonomous Vehicle and Drone Collaboration

This paper addresses the output synchronization problem in heterogeneous leader-follower multi-agent systems with unknown follower dynamics. A data-driven control strategy is proposed, wherein a tracking controller is designed based on finite-horizon input-output data. The data is further employed to solve the output regulator equation, thereby facilitating synchronization of the followers’ outputs with that of the leader. To account for bounded noise interference, the matrix S-lemma is employed to reformulate the synchronization conditions into a tractable convex optimization problem. Furthermore, it is analytically shown that the synchronization error remains bounded and is explicitly constrained by the solution of the optimization problem. The proposed methodology is substantiated through comprehensive numerical simulations and experimental validations, demonstrating its effectiveness and robustness.