期刊文献+

多智能体系统中具有先验知识的Q学习算法 被引量:21

Q-learning with prior knowledge in multi-agent systems
原文传递
导出
摘要 为了提高多智能体系统中的典型的强化学习——Q学习的学习效率和收敛速度,充分利用环境信息和相关的专家经验,提出了具有先验知识的Q学习算法。利用模糊综合决策方法处理专家经验和环境信息得到Q学习的先验知识,对Q学习的初始状态进行优化。典型的足球机器人系统中的仿真实验结果表明:该算法使学习过程建立在较好的学习基础上,从而更快地趋近于最优状态,其学习效率和收敛速度明显优于普通的Q学习。 Reinforcement Learning (RL) is an important branch of machine learning and it is unsupervised without specific signals. The learning process adjusts its actions according to external signals from interactions with the environment as a result, the system learning speed is relatively slow. Q-learning is a typical RL method with a slow convergence speed especially as the scales of the state space and the action space increase. An improved Q-learning method using prior knowledge uses fuzzy integrated decision making to process expert knowledge, which optimizes the initial states to give a better learning foundation. Test results on the Robot Soccer system show that the improved Q learning method has a higher learning efficiency and convergence speed.
出处 《清华大学学报(自然科学版)》 EI CAS CSCD 北大核心 2005年第7期981-984,共4页 Journal of Tsinghua University(Science and Technology)
基金 山东省自然科学基金资助项目(Y2002G18)
关键词 机器学习 Q学习 模糊综合决策 多智能体系统 machine learning Q-learning fuzzy integrated decision-making multi-agent system
  • 相关文献

参考文献7

二级参考文献25

  • 1庄晓东,孟庆春,魏天滨,王旭柱,谭锐,李筱菁.Robot path planning in dynamic environment based on reinforcement learning[J].Journal of Harbin Institute of Technology(New Series),2001,8(3):253-255. 被引量:3
  • 2田锡唐.焊接结构[M].北京:机械工业出版社,1981..
  • 3王元良.焊接变形与应用[M].成都:西南交通大学出版社,1990..
  • 4候正礼 习振中.工程项目评估模型系统理论与方法[M].科学出版社,1991.96-112.
  • 5刘砚田 郭志恭 等.工程经济[M].西安:西安交通大学出版社,1998.126-148.
  • 6Watkins C J C H. Learning from Delayed Rewards:[Ph.D.thesis]. Cambridge University, 1989.
  • 7Watkins C J C H. Dayan P. Technical not:Q-learning. Machine Learning, 1992,8:279~292.
  • 8Ohashi T ,et al. State transition rate based reinforcement learning Systems, Man, and Cybernetics. In: 2000 IEEE Intl. Cord.Volume: 1, 2000. 236~241.
  • 9Yamagnchi T,et al. Propagating learned behaviors from a virtual agent to a physical robot in reinforcement learnins, In..Proe. IEEE Int. Conf. on Evolutionary Computation, 1996. 855~859.
  • 10Yamagnchi T,et al. Reinforcement learning for a real robot in a real environment. In: European Conf. on Artificial Intelligence,Aug. 1996. 694~698.

共引文献101

同被引文献182

引证文献21

二级引证文献92

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部