arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.21680cs.LGcs.MA
收藏

基于分层Lead Critic的多智能体强化学习

Hierarchical Lead Critic based Multi-Agent Reinforcement Learning

Fraunhofer Institute IVI(弗劳恩霍夫研究所)

David Eckel, Henri Meeß

AI总结:

本文提出分层Lead Critic框架,通过多层级视角提升多智能体强化学习的样本效率和鲁棒性。

摘要:

合作多智能体强化学习(MARL)解决需要多个智能体协调的复杂任务,但通常局限于局部(独立学习)或全局(集中学习)视角。本文介绍了一种新的顺序训练方案和MARL架构,该架构从不同层级的多个视角进行学习。我们提出了受团队结构中自然出现分布启发的分层Lead Critic(HLC),其中高层目标与低层执行相结合。HLC证明引入多个层级,利用局部和全局视角,可以带来更高的样本效率和鲁棒策略性能。在合作、非通信和部分可观测的MARL基准测试中,实验结果表明HLC优于单一层级基线,并在智能体数量和难度增加时表现出鲁棒性。

英文摘要:

Cooperative Multi-Agent Reinforcement Learning (MARL) solves complex tasks that require coordination from multiple agents, but is often limited to either local (independent learning) or global (centralized learning) perspectives. In this paper, we introduce a novel sequential training scheme and MARL architecture, which learns from multiple perspectives on different hierarchy levels. We propose the Hierarchical Lead Critic (HLC) - inspired by natural emerging distributions in team structures, where following high-level objectives combines with low-level execution. HLC demonstrates that introducing multiple hierarchies, leveraging local and global perspectives, can lead to improved performance with high sample efficiency and robust policies. Experimental results conducted on cooperative, non-communicative, and partially observable MARL benchmarks demonstrate that HLC outperforms single hierarchy baselines and scales robustly with increasing amounts of agents and difficulty.

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。

↑