Although large language models (LLMs) have succeeded in natural language understanding, there are still many challenges in converting natural language instructions into understandable and executable action plans for UAVs while generalizing across different missions. This paper introduces an LLM-based UAV swarm task planning method that improves traditional multi-UAV task planning by utilizing the excellent comprehension capabilities of LLMs. First, the methodology uses elements such as AirSim simulation environment information, task types, and UAV behavior as the basis for UAV mission planning by designing Prompts for integration with a manually created library of high-level UAV functions. Then, we let the LLM take on the role of a planner to plan the task commands from the human operator into a sequence of action, i.e., generating a multi-step task plan that can adapt to different mission requirements such as UAV swarm formation control, and coverage reconnaissance. After that, we construct autoregressive detection and accurate mapping of action commands-UAV functions to avoid the drawbacks of unpredictability and insecurity by using the LLM-generated control codes directly without a closed loop. Finally, the simulation experiments of different scenarios demonstrate that the proposed method has higher accuracy and better stability.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    LLMSTP: Empowering Swarm Task Planning with Large Language Models


    Weitere Titelangaben:

    Lect. Notes Electrical Eng.


    Beteiligte:
    Liu, Lianqing (Herausgeber:in) / Niu, Yifeng (Herausgeber:in) / Fu, Wenxing (Herausgeber:in) / Qu, Yi (Herausgeber:in) / Yu, Hongbo (Autor:in) / Wang, Chang (Autor:in) / Wu, Lizhen (Autor:in) / Liu, Yuhao (Autor:in) / Niu, Yifeng (Autor:in)

    Kongress:

    International Conference on Autonomous Unmanned Systems ; 2024 ; Shenyang, China September 19, 2024 - September 21, 2024



    Erscheinungsdatum :

    09.04.2025


    Format / Umfang :

    11 pages





    Medientyp :

    Aufsatz/Kapitel (Buch)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    Empowering the planning controller

    Graham, R. / Marsden, A. / Dowling, F. et al. | IEEE | 1999


    Experience Adapter: Adapting Pre-trained Language Models for Continual Task Planning

    Zhang, Jiatao / Liao, Jianfeng / Hu, Tuocheng et al. | Springer Verlag | 2023


    Experience Adapter: Adapting Pre-trained Language Models for Continual Task Planning

    Zhang, Jiatao / Liao, Jianfeng / Hu, Tuocheng et al. | TIBKAT | 2023


    UAV Visual Path Planning Using Large Language Models

    Samma, Hussein / El-Ferik, Sami | Elsevier | 2025

    Freier Zugriff