Q学习在RoboCup前场进攻动作决策中的应用  

Application of Q-Learning in local attacking decision

作　　者：章惠龙[1] 李龙澍[1] 

机构地区：[1]安徽大学计算机科学与技术学院

出　　处：《计算机工程与应用》2013年第7期240-242,共3页Computer Engineering and Applications

基　　金：安徽省自然科学基金(No.090412054)

摘　　要：RoboCup是世界上规模最大的机器人足球大赛,包括软件仿真与硬件实体两类项目的比赛。RoboCup仿真2D作为软件仿真项目的重要组成部分,成为研究人工智能和多Agent智能体协作的优秀实验平台。将Q学习应用到RoboCup仿真2D比赛的前场进攻动作决策中,通过引入区域划分,基于区域划分的奖惩函数和对真人足球赛中动作决策的模拟,在经过大量周期的学习训练后,使Agent能够进行自主动作决策,从而加强了多Agent的前场进攻实力。RoboCup （Robot World Cup） is the largest scale robot soccer game, including software simulation and hardware enti- ties from two categories project competition. As an important part of software simulation project, RoboCup simulation 2D has become an outstanding experiment platform in which artificial intelligence and multi-agent cooperation are studied. This paper applies the Q-Learning to RoboCup simulation 2D match local attacking decision, through the introduction of zoning, incentive functions based zoning and decision making for real soccer game action simulation, after training a large number of cycles of leaming, making the Agent do the independent action decision, thereby strengthening the multi-agent attacking strength.

关 键 词：Q学习 ROBOCUP 多智能体协作 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]