城市群一体化与政府互动的多智能体模拟

Multi-Agent Simulation Research on Urban Agglomeration Integrationand Local Government Interaction

以城市群一体化与多政府互动为例,通过概念模型(因果关系图)、数学模型(局部规则设计)和计算机模型(Netlogo软件编程)的完整建模过程构建一个多主体仿真模型,模拟城市群多政府行为互动与态度演化的动态机理。研究表明:不论初始状态如何,群体互动结果总是由波动趋向于稳定,决策者应当有预见性、积极主动地采取引导调控措施,避免演化停滞于"非理想均衡";行政措施的引导效果比激励措施更加有效,多样化的调控手段可向行政措施适当倾斜;一体化偏好的初始状态对未来演化结果没有决定性影响。上层政府或协调机构应当全面、系统地考虑群体互动过程中的重要环节和关键因素,并有针对性地采取适当激励措施和行政措施,推进城市群一体化进程。

To analyze urban agglomeration integration and multi-government interaction,a multi-agent simulation model is built through conceptual model(causal reasoning chart),mathematic model(local rule design) and computer model(Netlogo programming),to simulate the dynamic process of the multi-governments’ behavioral interaction and attitude evolution.The experiments reveal that group interactions always tend to be stable,and decision-makers should adopt control measures actively and prematurely to avoid nonideal equilibrium;that the effects of administration measures exceed motivation measures,and the system of control measures might lean to administration measures aptly;and that the initial state of integration preferences have no determining influence on evolution results.Upper governments or coordination agencies should comprehensively consider the key elements in group interacting process,and accordingly adopt motivation and administration measures to advance the urban agglomeration integrating progress.