CHMAS: A Coupled Hierarchical Framework for Multi-Agent Reinforcement Learning
CHMAS:一种用于多智能体强化学习的耦合分层框架
专题命中 多智能体 :agent(title,abstract);multi-agent(title,abstract);planning(abstract)
AI总结 本文针对多智能体强化学习跨时间尺度平衡协调与执行的挑战,提出CHMAS框架,将决策分解为集中战略规划与分布式战术执行,有双向信息流和反馈机制,还给出异步更新协议,经理论分析与实验验证,该框架有效。
Comments Presented at the 2026 American Control Conference (ACC 2026). This version corrects the author names in Reference 1. The technical content, results, and conclusions are unchanged