Fuzzy Q-learning in continuous state and action space

被引：4
|
作者：
Xu M.-L. [1 ,2 ]
Xu W.-B. [2 ]
机构：
[1] Department of Electronic Information Engineering, Wuxi City College of Vocational Technology
[2] School of Information Technology, Jiangnan University
来源：
Journal of China Universities of Posts and Telecommunications | 2010年 / 17卷 / 04期
基金：
中国国家自然科学基金;
关键词：
adaptation; continuous; FIS; Q-learning;
D O I：
10.1016/S1005-8885(09)60495-7
中图分类号：
学科分类号：
摘要：
An adaptive fuzzy Q-learning (AFQL) based on fuzzy inference systems (FIS) is proposed. The FIS realized by a normalized radial basis function (NRBF) neural network is used to approach Q-value function, whose input is composed of state and action. The rules of FIS are created incrementally according to the novelty of each element of the pair of state-action. Moreover the premise part and consequent part of the FIS are updated using extended Kalman filter (EKF). The action that impacts on environment is the one with maximum output of FIS in the current state and generated through optimization method. Simulation results in the wall-following task of mobile robots and the inverted pendulum balancing problem demonstrate that the superiority and applicability of the proposed AFQL method. © 2010 The Journal of China Universities of Posts and Telecommunications.
引用
收藏
页码：100 / 109
页数：9
相关论文
共 50 条
  • [1] Fuzzy Q-learning in continuous state and action space
    XU Ming-liang1
    TheJournalofChinaUniversitiesofPostsandTelecommunications, 2010, 17 (04) : 100 - 109
  • [2] Reinforcement distribution in continuous state action space fuzzy Q-learning: A novel approach
    Bonarini, A
    Montrone, F
    Restelli, M
    FUZZY LOGIC AND APPLICATIONS, 2006, 3849 : 40 - 45
  • [3] Q-learning in continuous state and action spaces
    Gaskett, C
    Wettergreen, D
    Zelinsky, A
    ADVANCED TOPICS IN ARTIFICIAL INTELLIGENCE, 1999, 1747 : 417 - 428
  • [4] State and Action Space Segmentation Algorithm in Q-learning
    Notsu, Akira
    Ichihashi, Hidetomo
    Honda, Katsuhiro
    2008 IEEE INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, VOLS 1-8, 2008, : 2384 - 2389
  • [5] Multi-objective fuzzy Q-learning to solve continuous state-action problems
    Asgharnia, Amirhossein
    Schwartz, Howard
    Atia, Mohamed
    NEUROCOMPUTING, 2023, 516 : 115 - 132
  • [6] Reduction of the dynamic state-space in Fuzzy Q-Learning
    Kovács, S
    Baranyi, N
    2004 IEEE INTERNATIONAL CONFERENCE ON FUZZY SYSTEMS, VOLS 1-3, PROCEEDINGS, 2004, : 1075 - 1080
  • [7] Continuous-Action Q-Learning
    José del R. Millán
    Daniele Posenato
    Eric Dedieu
    Machine Learning, 2002, 49 : 247 - 265
  • [8] Continuous-action Q-learning
    Millán, JDR
    Posenato, D
    Dedieu, E
    MACHINE LEARNING, 2002, 49 (2-3) : 247 - 265
  • [9] Q-learning in Continuous State-Action Space with Redundant Dimensions by Using a Selective Desensitization Neural Network
    Kobayashi, Takaaki
    Shibuya, Takeshi
    Morita, Masahiko
    2014 JOINT 7TH INTERNATIONAL CONFERENCE ON SOFT COMPUTING AND INTELLIGENT SYSTEMS (SCIS) AND 15TH INTERNATIONAL SYMPOSIUM ON ADVANCED INTELLIGENT SYSTEMS (ISIS), 2014, : 801 - 806
  • [10] Continuous valued Q-learning method able to incrementally refine state space
    Takeda, M
    Nakamura, T
    Ogasawara, T
    IROS 2001: PROCEEDINGS OF THE 2001 IEEE/RJS INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS, VOLS 1-4: EXPANDING THE SOCIETAL ROLE OF ROBOTICS IN THE NEXT MILLENNIUM, 2001, : 265 - 271
12345