基于一般和随机对策论框架下的多智能体学习  

MULTI-AGENT LEARNING BASED ON GENERAL-SUM STOCHASTIC GAMES

作　　者：欧海涛[1] 张卫东[1] 许晓鸣[1] 

机构地区：[1]上海交通大学自动化系,上海200030

出　　处：《自动化学报》2002年第3期423-426,共4页Acta Automatica Sinica

基　　金：国家自然科学基金 ( 6 0 1 74 0 38)资助

摘　　要：将 Q- learning从单智能体框架上扩展到非合作的多智能体框架上 ,建立了在一般和随机对策框架下的多智能体理论框架和学习算法 ,提出了以 Nash平衡点作为学习目标 .给出了对策结构的约束条件 ,并证明了在此约束条件下算法的收敛性 。Q -learning from original single-agent framework is extended to non-cooperative multi-agent framework, and the theoretic framework of multi-agent learning is proposed under general-sum stochastic games with Nash equilibrium point as learning objective. We introduce a multi-agent Q -learning algorithm and prove its convergence under certain restriction, which is very important for the study and application of multi-agent system.

关 键 词：随机对策论 多智能体学习 学习算法 强化学习 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]