摘要
针对机器人比赛时局势的动态变化给机器人对抗决策博弈局面带来的不完全性问题,提出了豪尔绍尼转换和贝叶斯均衡相融合的不完全信息博弈算法,该算法克服了博弈局势中对未知信息的盲目"猜测".以机器人足球比赛时的数据为背景建立不完全信息博弈模型,研究机器人的决策对抗系统.仿真结果表明,不完全信息博弈算法可以使得机器人进行较优策略的选择,从而进一步提高机器人在比赛中的自主性和智能性.
While considering the incomplete game situation of decision-making in robot competition caused by the dynamic changes of the situation,games of incomplete information fusing Harsanyi conversion and Bayesian equilibria were proposed.This enabled the elimination of blind guessing of unknown information in game circumstances.Based on the robot soccer competition,a game model of incomplete information was established and a decision-making robot system was researched.The simulation results showed that the incomplete information game algorithm can help robots to choose better strategies,further improving the independence and intelligence of robots in competition.
出处
《智能系统学报》
2011年第2期147-151,共5页
CAAI Transactions on Intelligent Systems
基金
陕西省自然科学基金资助项目(2009JM8002)
关键词
不完全信息博弈
贝叶斯均衡
豪尔绍尼转换
对抗决策
机器人比赛
incomplete information game
Bayesian equilibria
Harsanyi conversion
decision-making
robot compctition
