SWARM-LLM:基于大语言模型的无人集群任务规划系统

doi:10.11896/jsjkx.241000038

计算机科学 ›› 2025, Vol. 52 ›› Issue (1): 72-79.doi: 10.11896/jsjkx.241000038

• 大语言模型技术研究及应用 • 上一篇下一篇

SWARM-LLM:基于大语言模型的无人集群任务规划系统

李婷婷¹, 王琪^1,2, 王嘉康^1,2, 徐勇军^1,2

1 中国科学院计算技术研究所北京 100190
2 中国科学院大学北京 100049

收稿日期:2024-10-09 修回日期:2024-11-26 出版日期:2025-01-15 发布日期:2025-01-09
通讯作者: 王琪(wangqi08@ict.ac.cn)
作者简介:(litingting@ict.ac.cn)

SWARM-LLM:An Unmanned Swarm Task Planning System Based on Large Language Models

LI Tingting¹, WANG Qi^1,2, WANG Jiakang^1,2, XU Yongjun^1,2

1 Institute of Computing Technology,Chinese Academy of Sciences,Beijing 100190,China
2 University of Chinese Academy of Sciences,Beijing 100049,China

Received:2024-10-09 Revised:2024-11-26 Online:2025-01-15 Published:2025-01-09
About author:LI Tingting,born in 1997,master.Her main research interests include group decision-making intelligence and large model theory and application.
WANG Qi,born in 1985,Ph.D,assiciate professor,Ph.D supervisor,is a sensior member of CCF(No.77141M).Her main research interests include intelligent wireless networks and LLM assisted decision making.

摘要/Abstract

摘要： 针对无人集群系统自主智能性不足、异构无人集群协同效率低、任务分配不均衡等问题,文中面向无人集群自主规划、高效协作、智能决策的需求,首先提出了一种新的基于大语言模型的无人集群任务规划系统框架(SWARM-LLM)。该框架利用大语言模型将高层次的任务指令转化为具体的智能无人集群任务规划方案,通过任务分解、任务分配、任务执行等多个阶段来实现无人集群协同任务。进一步地,设计了一套适用于无人集群规划的提示工程方法-规划链(Planning Chain,PC),用来指导和优化上述各阶段的实施。最终,在无人集群仿真环境(AirSim)中构建了不同类别和复杂度的任务,并进行了评估实验。与其他基于优化算法和机器学习的算法相比,实验结果证明了SWARM-LLM框架的有效性,并在任务成功率上展现出了显著的优势,平均性能提升了47.8%。

关键词: 任务规划, 无人集群, 大语言模型, 协同策略, 智能决策

Abstract: In response to the issues of insufficient autonomous intelligence in unmanned cluster systems,low collaborative efficiency of heterogeneous unmanned clusters,and unbalanced task allocation,this paper first proposes a new unmanned cluster task planning framework(SWARM-LLM) based on large language models to meet the needs of unmanned swarm systems for autonomous planning,efficient collaboration,and intelligent decision-making.This framework leverages large language models to transform high-level task instructions into specific intelligent unmanned cluster task planning solutions,achieving collaborative tasks of unmanned clusters through multiple stages such as task decomposition,task allocation,and task execution.Furthermore,this paper designs a prompt engineering method specifically suited for unmanned swarm planning,called the planning chain(PC),to guide and optimize the implementation of these stages.Finally,we construct tasks of various categories and complexities in an unmanned swarm simulation environment(AirSim) and conduct evaluation experiments.Compared with other algorithms based on optimization and machine learning,experimental results demonstrate the effectiveness of the SWARM-LLM framework,showing a significant advantage in task success rates,with an average performance improvement of 47.8%.

Key words: Task planning, Unmanned swarms, Large language models, Collaborative strategies, Intelligent decision-making

中图分类号:

TP181

李婷婷, 王琪, 王嘉康, 徐勇军. SWARM-LLM:基于大语言模型的无人集群任务规划系统[J]. 计算机科学, 2025, 52(1): 72-79. https://doi.org/10.11896/jsjkx.241000038

LI Tingting, WANG Qi, WANG Jiakang, XU Yongjun. SWARM-LLM:An Unmanned Swarm Task Planning System Based on Large Language Models[J]. Computer Science, 2025, 52(1): 72-79. https://doi.org/10.11896/jsjkx.241000038

参考文献

[1]JIANG B T,WEN G H,ZHOU J L,et al.Research Status and Prospects of Cross-Domain Collaborative Technology for Intelligent Unmanned Swarm Systems [J].Chinese Engineering Science,2024,26(1):117-126.
[2]WANG X,WANG H,SUN Q,et al.Air-ground coordinated unmanned swarm systems:A multitasking framework for control design[J].ISA Transactions,2024,145:315-329.
[3]ZHAO W,ZHOU K,TANG T,et al.A Survey of Large Language Models[J].arXiv:2303.18223,2023.
[4]RIBEIRO M T,WU T,GUESTRIN C,et al.Beyond Accuracy:Behavioral Testing of NLP models with CheckList[J].Computing Research Repository 2020(2020):4902-4912.
[5]LIANG J,HUANG W,XIA F,et al.Code as policies:Language model programs for embodied control[C]//2023 IEEE International Conference on Robotics and Automation(ICRA).IEEE,2023:9493-9500.
[6]QI X,LI B,FAN Y,et al.A survey of mission planning on UAVs systems based on multiple constraints[J].CAAI Transactions on Intelligent Systems,2020,15(2):204-217.
[7]WANG Q,LI T,XU Y,et al.How to prevent malicious use of intelligent unmanned swarms?[J].The Innovation,2023,4(2):209-211.
[8]CHEN J C,LING F Y,ZHANG Y,et al.Coverage path planning of heterogeneous unmanned aerial vehicles based on ant colony system[J].Swarm and Evolutionary Computation,2022,69:101005.
[9]XIA Y S,SHI J M.Vehicle-Aircraft Collaborative Multi-Area Coverage Reconnaissance Path Planning Method [J].Journal of Command and Control,2020,6(4):372-380.
[10]ZHANG T,LI C,MA D,et al.An optimal task management and control scheme for military operations with dynamic game stra-tegy[J].Aerospace Science and Technology,2021,115:106815.
[11]TANG J,LIANG Y,LI K.Dynamic Scene Path Planning of UAVs Based on Deep Reinforcement Learning[J].Drones,2024,8(2):60.
[12]LIU C,MA X,GAO X,et al.Distributed energy-efficient multi-UAV navigation for long-term communication cov-erage by deep reinforcement learning[J].IEEE Transac-tions on Mobile Computing,2020,19(6):1274-1285.
[13]HE L,AOUF N,SONG B.Explainable Deep ReinforcementLearning for UAV autonomous path planning[J].Aero-space Science and Technology,2021,118:107052.
[14]FU D,LI X,WEN L,et al.Drive like a human:Rethinking autonomous driving with large language mod-els[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.2024:910-919.
[15]CHEN L,SINAVSKI O,HÜNERMANN J,et al.Driving with llms:Fusing object-level vector modality for explainable autonomous driving[J].arXiv:2310.01957,2023.
[16]VEMPRALA S,BONATTI R,BUCKER A,et al.Chatgpt for robot-ics:Design principles and model abilities[J].arXiv:2306.17582,2023.
[17]SHAH S,DEY D,LOVETT C,et al.Airsim:High-fidelity vi-sual and physical simu-lation for autonomous vehicles[C]//Field and Service Robotics:Results of the 11th International Confe-rence.2017:621-635.
[18]TAZIR M L,MANCAS M,DUTOIT T.From Words to Flight:Integrating OpenAI ChatGPT with PX4/Gazebo for Natural Language-Based Drone Control[C]//International Workshop on Computer Science and Engineering.2023.
[19]CHEN G,YU X,ZHONG L.TypeFly:Flying Drones withLarge Language Model[J].arXiv:2312.14950,2023.
[20]YE Q,AXMED M,PRYZANT R,et al.Prompt engineering a prompt engineer[J].arXiv:2311.05661,2023.
[21]KANNAN S S,VENKATESH V L N,MIN B C.SWARM-llm:SWARM multi-agent robot task planning using large lan-guage models[J].arXiv:2309.10062,2023.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

SWARM-LLM:基于大语言模型的无人集群任务规划系统

SWARM-LLM:An Unmanned Swarm Task Planning System Based on Large Language Models

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 0

Metrics

本文评价

推荐阅读 0