A Hierarchical Reinforcement Learning Based Approach for Multi-robot Cooperation in Unknown Environments

  • Yifan Cai
  • Simon X. Yang
  • Xin Xu
  • Gauri S. Mittal
Part of the Advances in Intelligent and Soft Computing book series (AINSC, volume 144)

Abstract

Reinforcement learning is a good method for multi-robot systems to handle tasks in unknown environments or with obscure models. MAXQ is a hierarchical reinforcement learning algorithm, which is limited by some inherent problems. In addition, much research has focused on the completion of the task, rather than the ability to deal with new tasks. In this paper, an improved MAXQ approach is adopted to tune the parameters of the cooperation rules. The proposed scheme is applied to target searching tasks by multi-robots. The simulation results demonstrate the effectiveness and efficiency of the proposed scheme.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Ming, G.F., Hua, S.: Course-scheduling algorithm of option-based hierarchical reinforcement learning. In: The 2nd International Workshop on Education Technology and Computer Science, ETCS 2010, Wuhan, Hubei, China, pp. 288–291 (March 2010)Google Scholar
  2. 2.
    Du, X., Li, Q., Han, J.: An analysis and hierarchical decomposition for hams. In: 2009 International Conference on Computational Science and Engineering, Vancouver, BC, Canada, pp. 1050–1054 (August 2009)Google Scholar
  3. 3.
    Mirzazadeh, F., Behsaz, B., Beigy, H.: A new learning algorithm for the maxq hierarchical reinforcement learning method. In: The 2007 International Conference on Information and Communication Technology, Dhaka, Bangladesh, pp. 105–108 (March 2007)Google Scholar
  4. 4.
    Cheng, X., Shen, J., Liu, H., Gu, O.: Multi-robot cooperation based on hierarchical reinforcement learning. In: The 7th International Conference on Computational Science, Beijing, China, pp. 90–97 (May 2007)Google Scholar
  5. 5.
    Juang, C.F., Hsu, C.H.: Reinforcement ant optimized fuzzy controller for mobile-robot wall-following control. IEEE Transactions on Industrial Electronics 56(10), 3931–3940 (2009)CrossRefGoogle Scholar
  6. 6.
    Jang, J.R., Sun, C., Mizutani, E.: Neuro-Fuzzy and Soft Computing: a computational approach to learning and machine intelligence. Prentice Hall, New York (1997)Google Scholar

Copyright information

© Springer-Verlag GmbH Berlin Heidelberg 2012

Authors and Affiliations

  • Yifan Cai
    • 1
  • Simon X. Yang
    • 1
  • Xin Xu
    • 2
  • Gauri S. Mittal
    • 1
  1. 1.School of EngineeringUniversity of GuelphGuelphCanada
  2. 2.College of Mechatronics and AutomationNational University of Defense TechnologyChangshaChina

Personalised recommendations