arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.06733cs.LGcs.AIcs.MA
收藏

成对不够:用于多智能体路径规划的超图神经网络

Pairwise is Not Enough: Hypergraph Neural Networks for Multi-Agent Pathfinding

University of Cambridge(剑桥大学);National Institute of Advanced Industrial Science and Technology(国家先进工业科学与技术研究院)

Rishabh Jain, Keisuke Okumura, Michael Amir, Pietro Lio, Amanda Prorok

AI总结:

本文提出HMAGAT超图多智能体注意力网络,通过超图注意力机制捕捉群体动态,解决多智能体路径规划中的注意力稀释问题,展现超图表示在复杂交互中的优势。

摘要:

多智能体路径规划(MAPF)是一个典型的多智能体协调问题,多个智能体需导航至各自目标且不发生碰撞。解决MAPF最优已知为NP难,导致采用学习方法以减轻在线计算负担。现有方法如图神经网络(GNN)通常仅限于智能体间的成对信息传递。然而,这种限制导致次优行为及关键问题,如注意力稀释,尤其在密集环境中群体协调最为关键。尽管高阶交互的重要性,现有方法未能充分探索。为解决这一表示瓶颈,我们引入HMAGAT(超图多智能体注意力网络),一种利用超图注意力机制的新型架构,以显式捕捉群体动态。实证表明,HMAGAT在学习MAPF求解器中达到新状态:尽管仅1M参数且训练数据少100倍,仍优于当前SoTA 85M参数模型。通过分析HMAGAT的注意力值,我们展示了超图表示如何缓解GNN中的注意力稀释问题,并捕捉成对方法无法处理的复杂交互。我们的结果表明,适当归纳偏置往往比训练数据大小或单纯参数数量对多智能体问题更为关键。

英文摘要:

Multi-Agent Path Finding (MAPF) is a representative multi-agent coordination problem, where multiple agents are required to navigate to their respective goals without collisions. Solving MAPF optimally is known to be NP-hard, leading to the adoption of learning-based approaches to alleviate the online computational burden. Prevailing approaches, such as Graph Neural Networks (GNNs), are typically constrained to pairwise message passing between agents. However, this limitation leads to suboptimal behaviours and critical issues, such as attention dilution, particularly in dense environments where group (i.e. beyond just two agents) coordination is most critical. Despite the importance of such higher-order interactions, existing approaches have not been able to fully explore them. To address this representational bottleneck, we introduce HMAGAT (Hypergraph Multi-Agent Attention Network), a novel architecture that leverages attentional mechanisms over directed hypergraphs to explicitly capture group dynamics. Empirically, HMAGAT establishes a new state-of-the-art among learning-based MAPF solvers: e.g., despite having just 1M parameters and being trained on 100$\times$ less data, it outperforms the current SoTA 85M parameter model. Through detailed analysis of HMAGAT's attention values, we demonstrate how hypergraph representations mitigate the attention dilution inherent in GNNs and capture complex interactions where pairwise methods fail. Our results illustrate that appropriate inductive biases are often more critical than the training data size or sheer parameter count for multi-agent problems.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑