中文摘要：

该文提出基于有限理性博弈最优反应动态的资源分配机制,使得网格用户在不需要掌握其他用户信息的情况下,通过不断地快速学习,进行策略调整而达到网格资源分配的稳定均衡,克服了资源分配博弈过程中网格用户完全理性的限制。仿真表明该机制可实现网格用户在有限理性下效用最大化的资源分配,并可校正用户的决策偏离。

英文摘要：

A grid resource allocation mechanism based on the best-response dynamics is introduced in this paper, with which the grid users study and adjust strategy through repeated game to achieve stable equilibrium without the completed information of the others. The performance result validate this mechanism is efficient to lead to an optimal grid resource allocation which maximums the utility of bounded rational grid user, and revise the defected strategy if the grid user take some inaccurate action.