| 一种模糊强化学习算法及其在RoboCup中的应用 |
| |
| 引用本文： | 高建清,王浩,于磊,方宝富.一种模糊强化学习算法及其在RoboCup中的应用[J].计算机工程与应用,2006,42(6):52-54. |
| |
| 作者姓名： | 高建清 王浩 于磊 方宝富 |
| |
| 作者单位： | 合肥工业大学计算机与信息学院,合肥,230009 |
| |
| 摘 要： | 传统的强化学习算法只能解决离散状态空间和动作空间的学习问题。论文提出一种模糊强化学习算法,通过模糊推理系统将连续的状态空间映射到连续的动作空间,然后通过学习得到一个完整的规则库。这个规则库为Agent的行为选择提供了先验知识,通过这个规则库可以实现动态规划。作者在RoboCup环境中验证了这个算法,实现了踢球策略的优化。
|
| 关 键 词： | 强化学习 模糊推理系统 模糊Q学习 RoboCup 踢球技术 |
| 文章编号： | 1002-8331-（2006）06-0052-03 |
| 收稿时间： | 2005-07 |
| 修稿时间： | 2005-07 |
A Fuzzy Reinforcement Learning Algorithm and Its Application in RoboCup Environment |
| |
| Gao Jianqing,Wang Hao,Yu Lei,Fang Baofu.A Fuzzy Reinforcement Learning Algorithm and Its Application in RoboCup Environment[J].Computer Engineering and Applications,2006,42(6):52-54. |
| |
| Authors: | Gao Jianqing Wang Hao Yu Lei Fang Baofu |
| |
| Affiliation: | Department of Computer and Engineering, Hefei University of Technology,Hefei 230009 |
| |
| Abstract: | Conventional reinforcement algorithms only deal with discrete state spaces and discrete action spaces.In this paper,we propose a fuzzy reinforcement algorithm,which map continuous state spaces to continuous action spaces by fuzzy inference system and then learn a rule base.The rule base provides prior knowledge for agent's action selection and dynamic programming.We confirm the algorithm in RoboCup environment and implement the optimization of kick skill. |
| |
| Keywords: | RoboCup |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
