项目名称: 复杂层级结构下博弈-控制系统策略分析和干预研究
项目编号: No.61304159
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 自动化技术、计算机技术
项目作者: 穆义芬
作者单位: 中国科学院数学与系统科学研究院
项目金额: 26万元
中文摘要: 本项目拟研究复杂层级结构下基于博弈的控制系统的策略分析和干预实现。这类系统广泛存在于社会经济系统中,是一类典型的复杂系统。由于系统中非对称博弈者之间存在层级结构和信息结构关系,个体之间的相互作用极为复杂,使得对其进行理论分析非常困难。本课题拟以经典博弈模型为背景和出发点,研究静态和动态环境下该系统的博弈均衡求解问题,考察层级结构对博弈均衡和系统运行的影响。以此为基础,本课题还将重点研究动态不确定性环境下为达到预定的系统特征对系统的干预方法,希望建立自适应学习和机制设计的框架、算法和分析结果。最后,本课题拟通过实验对理论结果进行检验修正,并尝试将结果用于实际问题。本课题的实现,既可以丰富控制论的内容,也可以从新的角度处理博弈论中的对称结构和静态环境等假设,并将为此类典型复杂系统的调控提供理论与技术支持。
中文关键词: 层级博弈;均衡;社会福利;激励能控性;最优化
英文摘要: This project will study the strategic analysis and intervention for the so-called game-based control systems. Such systems exist widely in the social and economic systems and are typical complex systems. Because of the hierarchical structures and information structures between the asymmetric players, the relationship between players is extremely complex and the general solution to this problem is very difficult to obtain. In this project, based on several classical game models, the equilibrium/solution of the game in both static and dynamic environment will be explored, and the systematic property under the solution will be investigated too. Then, the intervention implemention will be studied in order to achieve certain systematic requirement and the framework, algorithm and analysis of adaptive learning and mechanism design will be developed for such systems under dynamic uncertain environment. The completeness of this project will enhance the content of control theory and handle the imperfect hypothesis in game theory such as symmetric structure and static environment from a new view point, and finally will provide theoretical support and technical guidance to real systems.
英文关键词: hierarchical games;equilibrium;social welfare;incentive controllability;optimization