Benchmarking the Full-Order Model Optimization Based Imitation in the Humanoid Robot Reinforcement Learning Walk

被引:0
|
作者
Chaikovskaya, Ekaterina [1 ]
Minashina, Inna [1 ]
Litvinenko, Vladimir [1 ]
Davydenko, Egor [1 ]
Makarov, Dmitry [1 ,2 ]
Danik, Yulia [1 ,2 ]
Gorbachev, Roman [1 ]
机构
[1] Moscow Inst Phys & Technol, Moscow, Russia
[2] Russian Acad Sci, Fed Res Ctr Comp Sci & Control, Moscow, Russia
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
When a gait of a bipedal robot is developed using deep reinforcement learning, reference trajectories may or may not be used. Each approach has its advantages and disadvantages, and the choice of method is up to the control developer. This paper investigates the effect of reference trajectories on locomotion learning and the resulting gaits. We implemented three gaits of a full-order anthropomorphic robot model with different reward imitation ratios, provided sim-to-sim control policy transfer, and compared the gaits in terms of robustness and energy efficiency. In addition, we conducted a qualitative analysis of the gaits by interviewing people, since our task was to create an appealing and natural gait for a humanoid robot. According to the results of the experiments, the most successful approach was the one in which the average value of rewards for imitation and adherence to command velocity per episode remained balanced throughout the training. The gait obtained with this method retains naturalness (median of 3.6 according to the user study) compared to the gait trained with imitation only (median of 4.0), while remaining robust close to the gait trained without reference trajectories.
引用
下载
收藏
页码:206 / 211
页数:6
相关论文
共 50 条
  • [1] Teaching Humanoid Robot Reaching Motion by Imitation and Reinforcement Learning
    Savevska, Kristina
    Ude, Ales
    ADVANCES IN SERVICE AND INDUSTRIAL ROBOTICS, RAAD 2023, 2023, 135 : 53 - 61
  • [2] Generalized Model Learning for Reinforcement Learning on a Humanoid Robot
    Hester, Todd
    Quinlan, Michael
    Stone, Peter
    2010 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA), 2010, : 2369 - 2374
  • [3] Teaching a humanoid robot to walk faster through Safe Reinforcement Learning
    Garcia, Javier
    Shafie, Diogo
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2020, 88
  • [4] Optimization of a Compact Model for the Compliant Humanoid Robot COMAN Using Reinforcement Learning
    Colasanto, Luca
    Kormushev, Petar
    Tsagarakis, Nikolaos
    Caldwell, Darwin G.
    CYBERNETICS AND INFORMATION TECHNOLOGIES, 2012, 12 (03) : 76 - 85
  • [5] Humanoid robot control based on reinforcement learning
    Iida, S. (iida@ics.nitech.ac.jp), IEEE Robotics and Automation Society; Nagoya University, Japan; City of Nagoya, Japan; Nagoya City Science Museum; Chubu Science and Technology Center (Institute of Electrical and Electronics Engineers Inc.):
  • [6] Humanoid robot control based on reinforcement learning
    Iida, S
    Kuwayama, K
    Kanoh, M
    Kato, S
    Kunitachi, T
    Itoh, H
    PROCEEDINGS OF THE 2004 INTERNATIONAL SYMPOSIUM ON MICRO-NANOMECHATRONICS AND HUMAN SCIENCE, 2004, : 353 - 358
  • [7] Learning full-body motions from monocular vision: Dynamic imitation in a humanoid robot
    Cole, Jeffrey B.
    Grimes, David B.
    Rao, Rajesh P. N.
    2007 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS, VOLS 1-9, 2007, : 246 - +
  • [8] Full-Body Postural Control of a Humanoid Robot with Both Imitation Learning and Skill Innovation
    Gonzalez-Fierro, Miguel
    Balaguer, Carlos
    Swann, Nicola
    Nanayakkara, Thrishantha
    INTERNATIONAL JOURNAL OF HUMANOID ROBOTICS, 2014, 11 (02)
  • [9] Imitation learning of a model predictive controller for real-time humanoid robot walking
    de A. Porto, Vítor G.B.
    Melo, Dicksiano C.
    Maximo, Marcos R.O.A.
    Afonso, Rubens J.M.
    Engineering Applications of Artificial Intelligence, 2025, 143
  • [10] Imitation Game: A Model-based and Imitation Learning Deep Reinforcement Learning Hybrid
    Veith, Eric Msp
    Logemann, Torben
    Berezin, Aleksandr
    Wellssow, Arlena
    Balduin, Stephan
    2024 12TH WORKSHOP ON MODELING AND SIMULATION OF CYBER-PHYSICAL ENERGY SYSTEMS, MSCPES, 2024,