分布式强化学习在RoboCup中的应用  

Application of the Distributed Reinforcement Learning in RoboCup

作　　者：张振文[1] 程显毅[1] 李明[1] 

机构地区：[1]江苏大学计算机与通信工程学院,江苏镇江212013

出　　处：《现代电子技术》2007年第4期44-46,共3页Modern Electronics Technique

摘　　要：强化学习理论由于其自学习性和自适应性的优点而得到了广泛关注。基于主Agent的概念改进了传统的群体强化学习算法,应用于机器人足球(Robocup)仿真实验中,取得了初步结果。In the field of artificial intelligence, the reinforcement learning theory is receiving more and more attention with the advantage of its self - learning and self - adaptability. We present an algorithm of group reinforcement learning based on the conception of main Agent and application it into RoboCup. We get the primary result by experiment.

关 键 词：强化学习 分布式强化学习 AGENT ROBOCUP 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]