From Gridworlds to Warehouses: Adapting Lightweight One-shot Multi-Agent Pathfinding for AGVs
从网格世界到仓库:为AGVs适应轻量级一次性多智能体路径规划
机构 * National Institute of Advanced Industrial Science and Technology (AIST)(日本国家先进工业科学与技术研究院) ; Keio University(庆应大学)
专题命中 多智能体 :agent(title,abstract);multi-agent(title,abstract);planning(abstract)
AI总结 本文提出多智能体仓库路径规划(MAWPF),针对差分驱动AGVs的运动特性,引入四条约束条件,改进传统MAPF算法,通过实验验证PP和LNS2在多智能体场景下的不足,PIBT类方法在可扩展性上表现更优。
Comments To be presented at IJCAI 2026