基于演化博弈论的处罚与奖励规范研究

Studying the effects of punishment and reward norms on the collective action of cooperation for solving the problems of social cooperation has been a frontier and hotspot research topic in recent years. At present, most of the research work mainly focuses on unveiling the working mechanisms of some punishment and reward norms for the evolution of cooperation, and has not yet involved how to use the norms of punishment and reward for solving the problems of social cooperation. Using some analytical tools such as evolutionary game theory, system dynamics theory, and control theory, this project will systematically study the effects of punishment and reward norms on the collective action of cooperation as well as the emergence and optimization problems of punishment and reward norms. The planned research contents include as follows: First, the project plans to study the working mechanisms of implicated punishment, committee punishment, and committee reward for the evolution of cooperation as well as the dynamic characteristics of these proposed norms, and it plans to compare these proposed norms with other frequently-used punishment and reward norms. Then, it plans to study the interaction effects among different punishment and reward norms and the emergence of advanced norms. Finally, it plans to study the optimization problem of the use of punishment and reward norms by constructing the objective functions according to realistic requirements. The project will enrich and expand the study of punishment and reward norms, and explore some theoretical basis of norm-setting for efficiently solving the cooperation problems in the real society.

研究处罚和奖励规范对群体合作行为的作用来解决社会合作难题是近年来的一个前沿热点研究课题。目前，大部分研究工作主要集中在揭示一些处罚与奖励规范对合作演化的作用机理上，还没有涉及到如何运用处罚与奖励规范来解决社会合作难题。通过运用演化博弈论、系统动力学理论和控制理论等理论分析工具，本项目将系统地研究处罚与奖励规范对群体合作行为的作用以及处罚与奖励规范的涌现与优化问题。拟研究的内容包括：首先研究连坐处罚和委员会处罚与奖励规范对合作演化的作用机理及其动力学特性，并和其他常用的处罚与奖励规范进行比较；然后研究各种处罚与奖励规范之间的相互作用和强规范的涌现情况；最后根据实际要求建立目标函数来研究处罚与奖励规范使用的优化问题。项目将丰富和拓展对处罚与奖励规范的研究，并为有效地解决现实社会中的合作难题探索出一些规范制定的理论依据。

本项目采用演化博弈论、系统动力学理论与方法、马尔科夫链、稳定性理论、基于个体的计算机仿真方法等工具较系统地研究了一些处罚与奖励规范对群体合作的作用以及处罚与奖励规范的涌现与优化问题。首先分别建立了具有监督的连坐处罚规范、基于监督与举报的规范、条件性处罚规范等的演化博弈模型，研究了这些规范对群体合作的影响作用，揭示了这些规范的动力学特性。然后分别研究了排斥策略与惩罚策略、排斥策略与奖励策略在无结构群体中的相互作用情况，发现排斥策略能够战胜相应的惩罚策略，排斥策略也能够战胜相应的奖励策略；并且计算机仿真发现在空间结构群体中排斥策略和合作策略能够形成同盟战胜背叛策略。最后提出了基于惩罚与排斥的切换策略，发现这种切换策略在演化的公共品博弈中能够和合作策略、背叛策略共存，具有较大的演化优势。项目研究加深理解了基于处罚与奖励的规范促进群体合作的作用效果及其演化规律与涌现机理，从而推动和完善了博弈群体中合作行为演化与涌现课题的研究。