🤖 AI Summary
This work addresses the challenges of safety and coordination in decentralized multi-agent systems operating under partial observability, limited communication, and dynamic interactions while collaboratively discovering and capturing randomly appearing targets. To this end, the authors propose a decentralized multi-agent reinforcement learning framework that integrates a graph attention network (GAT) encoder to fuse local observations with neighboring agents’ communicated information, incorporates a safety filter to ensure trajectory safety, and introduces a structured reward function that encourages orthogonality among communication vectors to enhance collaborative efficiency. Experimental results based on the MAPPO algorithm demonstrate that the proposed method outperforms baseline approaches in task success rate, obstacle avoidance, communication decorrelation, and training stability. Ablation studies further confirm the effectiveness of the designed reward mechanism.
📝 Abstract
This paper introduces a decentralized multi-agent reinforcement learning framework enabling structurally heterogeneous teams of agents to jointly discover and acquire randomly located targets in environments characterized by partial observability, communication constraints, and dynamic interactions. Each agent's policy is trained with the Multi-Agent Proximal Policy Optimization algorithm and employs a Graph Attention Network encoder that integrates simulated range-sensing data with communication embeddings exchanged among neighboring agents, enabling context-aware decision-making from both local sensing and relational information. In particular, this work introduces a unified framework that integrates graph-based communication and trajectory-aware safety through safety filters. The architecture is supported by a structured reward formulation designed to encourage effective target discovery and acquisition, collision avoidance, and de-correlation between the agents'communication vectors by promoting informational orthogonality. The effectiveness of the proposed reward function is demonstrated through a comprehensive ablation study. Moreover, simulation results demonstrate safe and stable task execution, confirming the framework's effectiveness.