Robust formation tracking of multi-agent systems via reinforcement learning-based hybrid impulsive control

Area of Science:

• Control Systems Engineering
• Artificial Intelligence
• Robotics

Background:

• Multi-agent systems require robust control for coordinated tasks like formation tracking.
• Existing continuous and impulsive control methods face limitations in dynamic and uncertain environments.

Purpose of the Study:

• To develop a robust H∞ formation tracking control strategy for multi-agent systems.
• To enhance system adaptability, corrective actions, and robustness using a hybrid impulsive approach.
• To introduce an adaptive optimization framework for control parameters and impulsive moments.

Main Methods:

• Employing a hybrid impulsive control approach combined with reinforcement learning.
• Utilizing the Razumikhin technique for robust stabilization analysis.
• Implementing feasible constraints for sustained H∞ performance.
• Introducing the Hierarchical Multi-agent Cooperative Reinforcement Learning (HMAC-RL) framework.

Main Results:

• The proposed hybrid impulsive control framework demonstrates enhanced adaptability and faster corrective actions.
• The HMAC-RL framework successfully optimizes control parameters and impulsive moments.
• Numerical simulations validate the theoretical analysis and the efficiency of the HMAC-RL framework.

Conclusions:

• The developed hybrid impulsive control strategy significantly improves the robustness of multi-agent formation tracking.
• The HMAC-RL framework offers an effective approach for adaptive optimization in complex control systems.
• This research advances the field of robust control for multi-agent systems in uncertain environments.