基于帮助博弈模型的社会合作演化机制研究

项目名称： 基于帮助博弈模型的社会合作演化机制研究

项目编号： No.71201129

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 管理科学与工程

项目作者： 于同奎

作者单位： 西南大学

项目金额： 19万元

中文摘要： 本课题研究如何在大量自私个体中促成社会整体合作。通过将社会看作一个由众多自适应个体构成的复杂演化系统，研究：（1）社会道德规范- - 尤其是惩恶扬善（即对合作者合作，对不合作者惩罚）的规范- - 对促进社会合作的作用；（2）好的社会规范如何在大量自私个体的交互中涌现；（3）当信誉机制不健全时，权利不对称（即惩罚者可以少受或免受惩罚）如何促成腐败型社会合作；（4）代表性惩罚（如警察）和对执法者的监督机制能否促成清廉型社会合作。拟采用的研究方法有：（1）种群动力学，将采取不同策略（策略可能是合作或者不合作的）的个体看成生态系统中相互竞争的种群，利用演化博弈论刻画不同种群规模的演化；（2）互动行为实验，招募受试者开展实验，在可控环境下获取真实数据；（3）多主体计算机仿真，针对个体自底向上建模，研究社会合作演化的基层运行机制。该研究对理解社会运行规律、促进和谐社会建设具有启发意义。

中文关键词： 间接互惠；社会规范；演化博弈论；互动行为实验；多主体计算机模拟

英文摘要： We will study the promotion of social cooperation in a society with a large population of self-interested individuals. By regarding the society as a complex evolving system with many self-adaptive agents, we will focus on: (1) the role of social norm, especially norms provoking punishment to noncooperators and cooperation to cooperators, to the evoluion of social cooperation; (2) how a good social norm can emerge from the interactions of a great number of selfish individuals; (3) how corruptive cooperation can be maintained in a society without repuation mechanism if there is power asymmetry which means that the punisher can be less or no punished when defecting; (4) representative punishers (e.g. police), mass surveillance to police and the evolution of cooperation. The research techniques to be used includes: (1)populaiton dynamics (evolutionary game theory), by considering agents with different strategies as different species who compete with each other in ecological system, we will investigate the fitness and populaiton size of these species; (2) interactive behavioral experiments, a experiment system will be designed based on platform Z-TREE, and subjects will be recruited to interact with each other in a virtual society, which can provide the direct and reliable individual behavioral pattern; (3) multi-ag

英文关键词： indirect reciprocity；social norm；evolutionary game theory；interactive behavioral experiment；multi-agent computer simulation