Learning-Based End-to-End Navigation for Planetary Rovers Considering Non-Geometric Hazards

被引:5
|
作者
Feng, Wenhao [1 ]
Ding, Liang [1 ]
Zhou, Ruyi [1 ]
Xu, Chongfu [1 ]
Yang, Huaiguang [1 ]
Gao, Haibo [1 ]
Liu, Guangjun [2 ]
Deng, Zongquan [1 ]
机构
[1] Harbin Inst Technol, State Key Lab Robot & Syst, Harbin 150080, Peoples R China
[2] Toronto Metropolitan Univ, Dept Aerosp Engn, Toronto, ON M5B 2K3, Canada
基金
中国国家自然科学基金;
关键词
Autonomous vehicle navigation; reinforcement learning; space robotics and automation; VISUAL NAVIGATION; MARS;
D O I
10.1109/LRA.2023.3281261
中图分类号
TP24 [机器人技术];
学科分类号
080202 ; 1405 ;
摘要
Autonomous navigation plays an increasingly crucial role in rover-based planetary missions. End-to-end navigation approaches developed upon deep reinforcement learning have enabled great adaptability in complex environments. However, most existing works focus on geometric obstacle avoidance thus have limited capability to cope with ubiquitous non-geometric hazards, such as sinkage and slippage. Autonomous navigation in unstructured harsh environments remains a great challenge requiring further in-depth study. In this letter, a DRL-based navigation method is proposed to autonomously guide a planetary rover towards goals via hazard-free paths with low wheel slip ratios. We introduce an end-to-end network architecture, in which the visual perception and the wheel-terrain interaction are fused to learn the representation of terrain mechanical properties implicitly and further facilitate policy learning for non-geometric hazard avoidance. Our approach outperforms baseline methods in simulation evaluation with superior avoidance capabilities against geometric obstacles and non-geometric hazards. Experiments conducted at a Mars emulation site suggest the successful deployment of our approach on a planetary rover prototype and the capacity of dealing with locomotion risks in real-world navigation tasks.
引用
收藏
页码:4084 / 4091
页数:8
相关论文
共 50 条
  • [31] End-to-End Learning-Based Obstacle Avoidance for Fixed-Wing UAVs
    Wang, Teng
    Xu, Zhao
    Hu, Jinwen
    Zhang, Haozhe
    Chen, Zhiwei
    PROCEEDINGS OF 2022 INTERNATIONAL CONFERENCE ON AUTONOMOUS UNMANNED SYSTEMS, ICAUS 2022, 2023, 1010 : 3342 - 3351
  • [32] Unsupervised Deep Learning-based End-to-end Network for Anomaly Detection and Localization
    Olimov, Bekhzod
    Subramanian, Barathi
    Kim, Jeonghong
    2022 THIRTEENTH INTERNATIONAL CONFERENCE ON UBIQUITOUS AND FUTURE NETWORKS (ICUFN), 2022, : 444 - 449
  • [33] Scene Understanding in Deep Learning-Based End-to-End Controllers for Autonomous Vehicles
    Yang, Shun
    Wang, Wenshuo
    Liu, Chang
    Deng, Weiwen
    IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS, 2019, 49 (01): : 53 - 63
  • [34] SiPhyR: An End-to-End Learning-Based Optimization Framework for Dynamic Grid Reconfiguration
    Haider, Rabab
    Annaswamy, Anuradha
    Dey, Biswadip
    Chakraborty, Amit
    IEEE TRANSACTIONS ON SMART GRID, 2025, 16 (02) : 1248 - 1260
  • [35] An End-to-End Learning-Based Metadata Management Approach for Distributed File Systems
    Gao, Yuanning
    Gao, Xiaofeng
    Zhang, Ruisi
    Chen, Guihai
    IEEE TRANSACTIONS ON COMPUTERS, 2022, 71 (05) : 1021 - 1034
  • [36] Deep Learning-Based End-to-End Design for OFDM Systems With Hardware Impairments
    Wu, Cheng-Yu
    Lin, Yu-Kai
    Wu, Chun-Kuan
    Lee, Chia-Han
    IEEE OPEN JOURNAL OF THE COMMUNICATIONS SOCIETY, 2023, 4 : 2468 - 2482
  • [37] SuperDriverAI: Towards Design and Implementation for End-to-End Learning-based Autonomous Driving
    Aoki, Shunsuke
    Yamamoto, Issei
    Shiotsuka, Daiki
    Inoue, Yuichi
    Tokuhiro, Kento
    Miwa, Keita
    2023 IEEE VEHICULAR NETWORKING CONFERENCE, VNC, 2023, : 195 - 198
  • [38] End-to-End Deep Learning for Autonomous Navigation of Mobile Robot
    Kim, Ye-Hoon
    Jang, Jun-Ik
    Yun, Sojung
    2018 IEEE INTERNATIONAL CONFERENCE ON CONSUMER ELECTRONICS (ICCE), 2018,
  • [39] Advanced Skills by Learning Locomotion and Local Navigation End-to-End
    Rudin, Nikita
    Hoeller, David
    Bjelonic, Marko
    Hutter, Marco
    2022 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), 2022, : 2497 - 2503
  • [40] End-to-End Learning of Joint Geometric and Probabilistic Constellation Shaping
    Aref, Vahid
    Chagnon, Mathieu
    2022 OPTICAL FIBER COMMUNICATIONS CONFERENCE AND EXHIBITION (OFC), 2022,