Daily Arxiv

This is a page that curates AI-related papers published worldwide.
All content here is summarized using Google Gemini and operated on a non-profit basis.
Copyright for each paper belongs to the authors and their institutions; please make sure to credit the source when sharing.

Task Assignment and Exploration Optimization for Low Altitude UAV Rescue via Generative AI Enhanced Multi-agent Reinforcement Learning

Created by
  • Haebom

Author

Xin Tang, Qian Chen, Wenjie Weng, Chao Jin, Zhang Liu, Jiacheng Wang, Geng Sun, Xiaohuan Li, Dusit Niyato

Outline

This paper focuses on the utilization of unmanned aerial vehicles (UAVs) with integrated artificial intelligence (AI) and ground robots (GERs) for emergency rescue operations in unknown environments. To overcome the computational limitations of a single UAV, we propose a cooperative framework that includes UAVs, GERs, and airships. The framework provides computational services for offloaded tasks by enabling resource pooling through UAV-GER (U2G) and UAV-Airship (U2A) links. We formulate the multi-objective problem of task assignment and navigation as a long-term dynamic optimization problem, aiming to minimize task completion time and energy consumption while ensuring stability. We transform the problem into a per-slot deterministic problem using Lyapunov optimization, and propose HG-MADDPG, which combines the Hungarian algorithm and GDM-based multi-agent deep deterministic policy gradient descent. Simulation results demonstrate significant improvements in offloading efficiency, latency, and system stability compared to baselines.

Takeaways, Limitations

•
Takeaways:
◦
Demonstrates that a collaborative framework integrating UAVs, GERs, and airships can improve the efficiency and safety of emergency rescue operations.
◦
We demonstrate that the proposed HG-MADDPG algorithm can be effectively applied to task assignment and navigation problems in multi-agent systems.
◦
We show that complex multi-objective optimization problems can be efficiently solved by utilizing the Lyapunov optimization technique.
•
Limitations:
◦
Lack of real-world validation of simulation results.
◦
Generalization performance evaluation for various environments and situations is required.
◦
Integration of airships may increase system complexity, which requires further study.
◦
Lack of analysis of the computational complexity of the proposed algorithm.
👍
Made with Slashpage