Multi-agent reinforcement learning for redundant robot control in task-space

被引:56
|
作者
Perrusquia, Adolfo [1 ]
Yu, Wen [1 ]
Li, Xiaoou [2 ]
机构
[1] CINVESTAV IPN Natl Polytech Inst, Dept Control Automat, Mexico City, DF, Mexico
[2] CINVESTAV IPN Natl Polytech Inst, Dept Comp, Mexico City, DF, Mexico
关键词
Multi-agent; Reinforcement learning; Redundant robot; TRACKING CONTROL; TRAJECTORY TRACKING; INVERSE KINEMATICS; MANIPULATORS;
D O I
10.1007/s13042-020-01167-7
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Task-space control needs the inverse kinematics solution or Jacobian matrix for the transformation from task space to joint space. However, they are not always available for redundant robots because there are more joint degrees-of-freedom than Cartesian degrees-of-freedom. Intelligent learning methods, such as neural networks (NN) and reinforcement learning (RL) can learn the inverse kinematics solution. However, NN needs big data and classical RL is not suitable for multi-link robots controlled in task space. In this paper, we propose a fully cooperative multi-agent reinforcement learning (MARL) to solve the kinematic problem of redundant robots. Each joint of the robot is regarded as one agent. The fully cooperative MARL uses a kinematic learning to avoid function approximators and large learning space. The convergence property of the proposed MARL is analyzed. The experimental results show that our MARL is much more better compared with the classic methods such as Jacobian-based methods and neural networks.
引用
收藏
页码:231 / 241
页数:11
相关论文
共 50 条
  • [1] Multi-agent reinforcement learning for redundant robot control in task-space
    Adolfo Perrusquía
    Wen Yu
    Xiaoou Li
    [J]. International Journal of Machine Learning and Cybernetics, 2021, 12 : 231 - 241
  • [2] Redundant Robot Control Using Multi Agent Reinforcement Learning
    Perrusquia, Adolfo
    Yu, Wen
    Li, Xiaoou
    [J]. 2020 IEEE 16TH INTERNATIONAL CONFERENCE ON AUTOMATION SCIENCE AND ENGINEERING (CASE), 2020, : 1650 - 1655
  • [3] Safe multi-agent reinforcement learning for multi-robot control
    Gu, Shangding
    Kuba, Jakub Grudzien
    Chen, Yuanpei
    Du, Yali
    Yang, Long
    Knoll, Alois
    Yang, Yaodong
    [J]. ARTIFICIAL INTELLIGENCE, 2023, 319
  • [4] A multi-agent reinforcement learning approach to robot soccer
    Yong Duan
    Bao Xia Cui
    Xin He Xu
    [J]. Artificial Intelligence Review, 2012, 38 : 193 - 211
  • [5] A multi-agent reinforcement learning approach to robot soccer
    Duan, Yong
    Cui, Bao Xia
    Xu, Xin He
    [J]. ARTIFICIAL INTELLIGENCE REVIEW, 2012, 38 (03) : 193 - 211
  • [6] Global task-space adaptive control of robot
    Li, Xiang
    Cheah, Chien Chern
    [J]. AUTOMATICA, 2013, 49 (01) : 58 - 69
  • [7] Multi-agent reinforcement learning for character control
    Li, Cheng
    Fussell, Levi
    Komura, Taku
    [J]. VISUAL COMPUTER, 2021, 37 (12): : 3115 - 3123
  • [8] Multi-Arm Robot Task Planning for Fruit Harvesting Using Multi-Agent Reinforcement Learning
    Li, Tao
    Xie, Feng
    Qiu, Quan
    Feng, Qingchun
    [J]. 2023 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS, IROS, 2023, : 4176 - 4183
  • [9] Multi-agent reinforcement learning for character control
    Cheng Li
    Levi Fussell
    Taku Komura
    [J]. The Visual Computer, 2021, 37 : 3115 - 3123
  • [10] FUZZY TASK-SPACE CONTROL OF A WELDING ROBOT
    Fateh, Mohammad M.
    [J]. INTERNATIONAL JOURNAL OF ROBOTICS & AUTOMATION, 2010, 25 (04): : 372 - 378