计算机科学

计算机科学 ›› 2025, Vol. 52 ›› Issue (3): 248-259.doi: 10.11896/jsjkx.241100068

• 人工智能 • 上一篇    下一篇

生成式任务网:基于大模型的自主任务规划与执行范式

黄雪芹, 张胜, 朱先强, 张千桢, 朱承   

  1. 国防科技大学信息系统工程全国重点实验室 长沙 410073
  • 收稿日期:2024-11-12 修回日期:2024-12-25 出版日期:2025-03-15 发布日期:2025-03-07
  • 通讯作者: 朱承(Zhucheng@nudt.edu.cn)
  • 作者简介:(hxq@nudt.edu.cn)
  • 基金资助:
    国防基础科研计划(WDZC20235250412);国防科技大学科研基金(ZK20-30)

Generative Task Network:New Paradigm for Autonomic Task Planning and Execution Based on LLM

HUANG Xueqin, ZHANG Sheng, ZHU Xianqiang, ZHANG Qianzhen, ZHU Cheng   

  1. National Key Laboratory of Information Systems Engineering,National University of Defense Technology,Changsha 410073,China
  • Received:2024-11-12 Revised:2024-12-25 Online:2025-03-15 Published:2025-03-07
  • About author:HUANG Xueqin,born in 1997,Ph.D student.His main research interests include large language model and intelligent task planning.
    ZHU Cheng,born in 1976,Ph.D,principal investigator.His main research interests include command and control,and intelligent decision-making.
  • Supported by:
    National Defense Basic Research Program(WDZC20235250412) and National University of Defense Technology Research Fund(ZK20-30).

PDF (PC)

233

摘要： 得益于生成式人工智能的发展,无人系统的智能规划技术将迎来新的变革。首先分析了传统智能任务规划范式在泛化性、可迁移性以及任务规划前后连贯性等方面的缺陷,针对性地提出了基于大模型的任务规划与执行新范式,即生成式任务网。该方法可以帮助无人系统实现任务自主发现、智能规划与自动执行,形成问题到解决的闭环,同时使无人系统的任务规划过程具备了可泛化和易迁移的优势。然后介绍了生成式任务网的内涵,并完成了它的要素定义和流程建模,进而设计了一个通用应用架构。最后以N航空公司航材库作为场景进行应用分析,有效提升了无人系统在仓库管理中的智能化和自动化水平。

关键词: 无人系统, 大模型, 任务规划, 任务执行, 生成式任务网

Abstract: Owing to the development of generative artificial intelligence,the intelligent planning technology of unmanned systems is set to undergo a new transformation.This paper first analyzes the shortcomings of traditional intelligent task planning paradigms in terms of generalization,transferability and coherence.In response,it proposes a new paradigm for task planning and execution based on large language models,namely generative task network.This method enables unmanned systems to autonomously discover tasks,intelligently plan,and automatically execute them,forming a closed-loop from problem to solution.It also endows the task planning process of unmanned systems with the advantages of generalization and ease of transfer.This paper then introduces the concept of the generative task network,defines its key elements and models its process,and subsequently designs a ge-neral application architecture.Finally,an application analysis is conducted taking the aviation materials warehouse of N Airlines as the scenario,effectively enhancing the intelligence and automation levels of unmanned systems in warehouse management.

Key words: Unmanned systems, Large language model, Task planning, Task execution, Generative task network

中图分类号: 

  • TP182
[1]OpenAI:Hello GPT-4o[EB/OL].(2024-05-14) [2024-09-01].https://openai.com/index/hello-gpt-4o/.
[2]ZENG T G A,XU B,WANG B,et al.ChatGLM:A Family of Large Language Models from GLM-130B to GLM-4 All Tools[J].arXiv:2406.12793,2024.
[3]FIORI L,DOSHI A,MARTINEZ E,et al.The Use of Un-manned Aerial Systems in Marine Mammal Research[J].Remote Sensing.2017,9(6):543.
[4]RAMESH P S,MURUGA L J J V.Evaluation of design criteria for mini unmanned aircraft systems(UAS) applications[J].Aircraft Engineering and Aerospace Technology,2022,94(3):327-335.
[5]SUN Z,WANG Q,CHEN L,et al.Unmanned Technology-Based Civil-Military Intelligent Logistics System:From Construction to Integration[J].Journal of Beijing Institute of Technology,2022,31(2):140-151.
[6]RIZK Y,AWAD M,TUNSTEL E W.Cooperative Heterogeneous Multi-Robot Systems:A Survey[J].ACM Computing Surveys,2020,52(2):1-31.
[7]WU X,XIAO B,CAO L,et al.Optimal Transport and ModelPredictive Control-based Simultaneous Task Assignment and Trajectory Planning for Unmanned System Swarm[J].Journal of Intelligent & Robotic Systems,2024(1):28.
[8]ZHU Z L,MA G B,HUANG P,et al.Establishment and Solution of Multi-satellite Imaging Scheduling Model Based on Task Decomposition[J].Spacecraft Engineering,2018(2):6-13.
[9]BI W,ZHANG M,CHEN H,et al.Cooperative task allocation method for air-sea heterogeneous unmanned system with an application to ocean environment information monitoring[J].Ocean Engineering,2024,309(Part 2):118496.
[10]LIU Q,HUANG K,LIU H,et al.Multi-Domain Collaborative Task Allocation and Conflict Resolution in Unmanned Systems under Complex Constraints[J].Unmanned Systems,2024,7(2):1-11.
[11]LIU H,LIAO Y,SHEN C S A J.Task Allocation of Heterogeneous Multi-Unmanned Systems Based on Improved Sheep Flock Optimization Algorithm[J].Future Internet,2024,16(4):1-20.
[12]SAMIEI A,ISMAIL S,SUN L.Cluster-based hungarian ap-proach to task allocation for unmanned aerial vehicles[C]//2019 IEEE National Aerospaceand Electronics Conference(NAECON).IEEE,2019:148-154.
[13]KARAMAN S,SHIMA T,FRAZZOLI E.Effective task assignment for complex uav operations using genetic algorithms[C]//AIAA Guidance,Navigation,and Control Conference.2013:6211.
[14]HAKSAR R N,SCHWAGER M.Distributed deep reinforce-ment learning for fighting forest fires with a network of aerial robots[C]//2018 IEEE/RSJ International Conference on Intelligent Robots andSystems(IROS).IEEE,2018:1067-1074.
[15]FU X,FENG P,GAO X.Swarm UAVs task and resource dynamic assignment algorithm based on task sequence mechanism[J].IEEE Access:Practical Innovations,Open Solutions,2019,7:41090-41100.
[16]WU Y,GOU J,JI H,et al.Hierarchical mission replanning for multiple UAV formations performing tasks in dynamic situation[J].Computer Communications,2023,200:132-148.
[17]SHAO T,ZHANG H,CHENG K,et al.Review of replanning in hierarchical task network[J].Systems Engineering and Electronics.2020,42(12):2833-2846.
[18]WARNER S M,ROYSET J O.Optimizing Surveillance Satellites for the Synthetic Theater Operations Research Model[J].Military Operations Research,2024,29(1):31-44.
[19]WEBER G,THOMAS J J,SAUCEDO J A M,et al.Correction to:Preface:Special Issue on Modeling,Simulation,and Optimization in Operational Research[J].Journal of the Operations Research Society of China,2024,12(2):548.
[20]GEORGIEVSKI I,AIELLO M.HTN planning:Overview,comparison,and beyond[J].Artificial Intelligence,2015,222(May):124-156.
[21]MORISSET B,GHALLAB M.Learning how to combine sensorymotor modalities for a robust behavior[C]//the International Seminar on Advances in Plan-Based Control of Robotic Agents.Heidelberg,Berlin:Springer,2002:157-178.
[22]SIRIN E,PARSIA B,WU D,et al.HTN planning for web service composition using SHOP2[J].Web Semantics:Science,Services and Agents on the World Wide Web,2004,1(4):377-396.
[23]VASWANI A,SHAZEER N,PARMAR N,et al.Attention isall you need[C]//Proceedings of the 31st International Confe-rence on Neural Information Processing Systems(NIPS’17).2017:6000-6010.
[24]BROWN T B,MANN B,RYDER N,et al.Language Models are Few-Shot Learners[J].arXiv:2005.14165,2020.
[25]TOUVRON H,LAVRIL T,IZACARD G,et al.LLaMA:Open and Efficient Foundation Language Models[J].arXiv:2302.13971,2023.
[26]TOUVRON H,MARTIN L,STONE K.Llama 2:Open Foundation and Fine-Tuned Chat Models[J].arXiv:2307.09288,2023.
[27]ZHANG Y,LI Y,CUI L,et al.Siren’s Song in the AI Ocean:A Survey on Hallucination in Large Language Models[J].arXiv:2309.01219,2023.
[28]PENG B,GALLEY M,HE P,et al.Check Your Facts and Try Again:Improving Large Language Models with External Knowledge and Automated Feedback[J].arXiv:2302.12813,2023.
[29]WANG L,MA C,FENG X,et al.A Survey on Large Language Model based Autonomous Agents[J].arXiv:2308.11432,2023.
[30]YAO S,ZHAO J,YU D,et al.ReAct:Synergizing Reasoningand Acting in Language Models[J].arXiv:2210.03629,2022.
[31]CHEN Z,ZHOU K,ZHANG B,et al.ChatCoT:Tool-Augmented Chain-of-Thought Reasoning on Chat-based Large Language Models[J].arXiv:2305.14323.2023.
[32]YAO S,YU D,ZHAO J,et al.Tree of thoughts:Deliberateproblem solving with large language models[J].arXiv:2305.10601,2023.
[33]ZHUANG Y,CHEN X,YU T,et al.Toolchain*:Efficient action space navigation in large language models with a* search[J].arXiv:2310.13227,2023.
[34]DAGAN G,KELLER F,LASCARIDES A.Dynamic planning with aLLM[J].arXiv:2308.06391,2023.
[35]SONG C H,WU J,WASHINGTON C,et al.LLM-planner:Few-shot grounded planning for embodied agents with large language models[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2023:2998-3009.
[36]CHEN W,SU Y,ZUO J,et al.Agentverse:Facilitating multi-agent collaboration and exploring emergent behaviors[C]//The Twelfth International Conference on Learning Representations.2023.
[37]TEAM X.XAgent:An autonomous agent for complex task solving [EB/OL].https://github.com/OpenBMB/XAgent.
[38]WU Q,BANSAL G,ZHANG J,et al.Autogen:Enabling next-gen llm applications via multi-agent conversation framework[J].arXiv:2308.08155,2023.
[39]PALANTIR.AIP for defense[EB/OL].https://www.palantir.com/platforms/aip/defense/.
[40]CORVEY W.Exploratory models of human-ai teams(emhat)[EB/OL].https://www.darpa.mil/program/exploratory-moldels-of-human-ai-teams
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发