城市群政府博弈与调控机制的多智能体系统建模——集成进化博弈理论与小世界网络模型

Multi-Agent System Modeling of Urban Agglomeration'Government Game and Control Mechanism——Aggregating Evolutionary Game and Small World Networks

以电子政务和移动政务环境下的城市群多政府博弈为例,基于多智能体系统的建模框架,展开系统运行机理分析、因果关系图设计和数学建模,基于进化博弈理论构建微观决策主体(即各城市政府决策者)的行为互动机制,并基于小世界网络模型构建宏观社会网络(即各决策者之间交互关系)的拓扑演化规则。且在模型构建中考虑了博弈主体之间的差异性和非对称性、收益受合作水平影响的动态性、以及设计了与惩罚措施配套的补偿机制。本文为基于智能体的城市群政府博弈模拟和政策实验研究奠定了模型基础。是复杂性科学理论和建模方法在政治学与公共行政领域的一次积极拓展,尝试了新的研究视角与方法集合。

Based on the multi - agent modeling framework, and through the process of system mecha- nism analysis, reasoning graph design, and mathematical modeling, a model for Urban Agglomeration＇ Government Game in the environment of E-government and M-government is built, constructing the be- havioral interaction mechanism of micro decision-making entities （namely urban government＇ decision- makers） based on the evolutionary game theory and building the topology evolving rule of macro social networks （namely interacting relationships among decision-makers） based on the small work networks model. Further, the difference and asymmetry among game agents, the dynamics of game payoff affected by levels of cooperation, and the new compensation mechanism accompanied with the punishment meas- ures are discussed and added into the model design. This paper founds the model basis for urban agglom-eration＇ s government game simulation and experimental policy study based on multi-agent system, which is a frontier expansion of complex science theories and modeling methods in politics and public administration, tries a new perspective and methodological assemble.