Online path planning for multiple unmanned aerial vehicle (multi-UAV) systems is considered a challenging task. It needs to ensure collision-free path planning in real-time, especially when the multi-UAV systems can become very crowded on certain occasions. In this paper, we presented a vision-based decentralized collision-avoidance policy for multi-UAV systems, which takes depth images and inertial measurements as sensory inputs and outputs UAV's steering commands. The policy is trained together with the latent representation of depth images using a policy gradient-based reinforcement learning algorithm and autoencoder in the multi-UAV threedimensional workspaces. Each UAV follows the same trained policy and acts independently to reach the goal without colliding or communicating with other UAVs. We validate our policy in various simulated scenarios. The experimental results show that our learned policy can guarantee fully autonomous collision-free navigation for multi-UAV in the three-dimensional workspaces with good robustness and scalability.


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Vision-based Distributed Multi-UAV Collision Avoidance via Deep Reinforcement Learning for Navigation


    Beteiligte:
    Huang, Huaxing (Autor:in) / Zhu, Guijie (Autor:in) / Fan, Zhun (Autor:in) / Zhai, Hao (Autor:in) / Cai, Yuwei (Autor:in) / Shi, Ze (Autor:in) / Dong, Zhaohui (Autor:in) / Hao, Zhifeng (Autor:in)

    Erscheinungsdatum :

    2022


    Anmerkungen:

    Submitted to IEEERSJ International Conference on Intelligent Robots and Systems (IROS) 2022


    Medientyp :

    Preprint


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    Vision-guided Collision Avoidance Through Deep Reinforcement Learning

    Song, Sirui / Zhang, Yuanhang / Qin, Xi et al. | IEEE | 2021



    Learning-Based Navigation and Collision Avoidance Through Reinforcement for UAVs

    Azzam, Rana / Chehadeh, Mohamad / Hay, Oussama Abdul et al. | IEEE | 2024

    Freier Zugriff

    Pedestrian Collision Avoidance Using Deep Reinforcement Learning

    Rafiei, Alireza / Fasakhodi, Amirhossein Oliaei / Hajati, Farshid | Springer Verlag | 2022


    Collision Avoidance Using Deep Learning-Based Monocular Vision

    Rill, Róbert-Adrian / Faragó, Kinga Bettina | Springer Verlag | 2021

    Freier Zugriff