arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.21670cs.ROcs.AIcs.MA
收藏

分层基于大语言模型的多智能体框架与提示优化用于多机器人任务规划

Hierarchical LLM-Based Multi-Agent Framework with Prompt Optimization for Multi-Robot Task Planning

Data Science Laboratories, NEC Corporation(日本电气株式会社数据科学实验室)

Tomoya Kawabe, Rin Takano

AI总结:

本文提出一种分层多智能体框架,结合提示优化提升多机器人任务规划的准确性和效率。

摘要:

多机器人任务规划需要将自然语言指令分解为可执行的动作以适应异构机器人团队。传统规划领域定义语言(PDDL)规划器提供严格保证,但难以处理模糊或长周期任务,而大语言模型(LLMs)可以解读指令并提出计划,但可能产生幻觉或不可行的动作。我们提出了一种分层多智能体基于LLM的规划器,结合提示优化:上层分解任务并分配给下层智能体,这些智能体生成由经典规划器解决的PDDL问题。当计划失败时,系统应用受TextGrad启发的文本梯度更新来优化每个智能体的提示,从而提高规划准确性。此外,元提示在相同层内的智能体之间学习和共享,使多智能体设置中的提示优化更高效。在MAT-THOR基准测试中,我们的规划器在复合任务上的成功率为0.95,在复杂任务上为0.84,在模糊任务上为0.60,分别比先前最先进的LaMMA-P提高了2、7和15个百分点。消融研究显示,分层结构、提示优化和元提示共享对总体成功率的贡献分别约为+59、+37和+4个百分点。

英文摘要:

Multi-robot task planning requires decomposing natural-language instructions into executable actions for heterogeneous robot teams. Conventional Planning Domain Definition Language (PDDL) planners provide rigorous guarantees but struggle to handle ambiguous or long-horizon missions, while large language models (LLMs) can interpret instructions and propose plans but may hallucinate or produce infeasible actions. We present a hierarchical multi-agent LLM-based planner with prompt optimization: an upper layer decomposes tasks and assigns them to lower-layer agents, which generate PDDL problems solved by a classical planner. When plans fail, the system applies TextGrad-inspired textual-gradient updates to optimize each agent's prompt and thereby improve planning accuracy. In addition, meta-prompts are learned and shared across agents within the same layer, enabling efficient prompt optimization in multi-agent settings. On the MAT-THOR benchmark, our planner achieves success rates of 0.95 on compound tasks, 0.84 on complex tasks, and 0.60 on vague tasks, improving over the previous state-of-the-art LaMMA-P by 2, 7, and 15 percentage points respectively. An ablation study shows that the hierarchical structure, prompt optimization, and meta-prompt sharing contribute roughly +59, +37, and +4 percentage points to the overall success rate.

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。

↑