Optimal cooperative output regulation with norm-based performance specifications.

Area of Science:

• Control Systems Engineering
• Multi-Agent Systems
• Optimization Theory

Background:

• Cooperative output regulation is crucial for coordinated multi-agent systems.
• Achieving explicit performance specifications with limited information is challenging.

Purpose of the Study:

• To develop an optimal cooperative output regulation strategy for linear heterogeneous multi-agent systems.
• To ensure both transient and steady-state performance of regulated errors while minimizing cost.
• To address scenarios where not all agents have direct access to the regulated error.

Main Methods:

• A hierarchical "observation-control" framework was constructed.
• A distributed edge-based observer was developed for state reproduction with prescribed transient performance.
• A data-driven actor-critic learning algorithm combined with prescribed performance control was employed for optimal control scheme derivation.

Main Results:

• The proposed method achieves explicit specifications on regulated errors indirectly through observation and auxiliary tracking errors.
• The multi-dimensional regulated errors converge to a predefined residual set within a specified time.
• The approach minimizes predefined cost without solving the nonlinear Hamilton-Jacobi-Bellman equation, reducing computational burden.

Conclusions:

• The developed hierarchical framework and data-driven control scheme effectively solve the optimal cooperative output regulation problem.
• Experimental results validate the proposed control scheme's effectiveness in ensuring performance and minimizing cost for multi-agent systems.