带有无界赔付函数的非零和随机对策折扣模型  

Nonzero-Sum Stochastic Games with Unbounded Discounted Payoff

作　　者：杨洁[1,2] 郭先平[2] 

机构地区：[1]北京信息科技大学理学院,北京100192 [2]中山大学数学与计算科学学院,广东广州510275

出　　处：《中山大学学报（自然科学版）》2008年第5期23-27,36,共6页Acta Scientiarum Naturalium Universitatis Sunyatseni

基　　金：国家自然科学基金资助项目(60574002)

摘　　要：讨论了赔付函数可能既无上界又无下界的离散时间可数状态非零和随机对策的折扣模型。在零和随机对策中常用的"漂移"和"连续-紧"性条件下,用Fan's不动点定理证明了Nash平衡点的存在性。Discrete time two-person nonzero-sum stochastic games with the discounted payoff criterion and a countable state space is studied, here the payoff functions might have neither upper nor lower bounds. The existence of Nash equilibria is proved in randomized stationary strategies by using the Fan＇s fixed-point theorem under the ＂drift＂ and the standard ＂continuous-compact＂ conditions, which are usually used in zero-sum stochastic games.

关 键 词：非零和随机对策 期望折扣赔付准则 NASH平衡点 可数状态空间 

分 类 号：O211[理学—概率论与数理统计]