Agent仿真技术在科技政策决策支持中的应用

摘要

目前采用基于Agent的建模与仿真(Agent- based Modeling and Simulation,ABMS)方法建立的人工社会模型大多是概念模型或者演示模型,还不能针对现实中的具体对象进行解释和分析.首先讨论采用ABMS方法建立人工社会模型的思路,然后采用ABMS方法针对具体的广州软件产业集群进行建模,并给出企业技术创新行为的仿真描述,最终建立的计算机仿真模型可以应用于分析政府税率、创新回报率、研发周期、社会网络和研发投入比例等因素对整个软件产业集群技术创新的影响,提供科技政策制定决策支持.%Most artificial society models which are built on agent - based modeling and simulation ( ABMS) methodology are only concept models or demos. These models can not be used to analyze and describe the real phenomena. Firstly, the paper discusses the idea of artificial society models on ABMS, and builds the concrete model of the software industrial cluster of Guangzhou by using ABMS methodology. The paper also gives the simulation description of enterprise innovation behavior in the model. The result shows that the model can be applied to policy making decision support. By the model, the effect of the change of government tax, return rate , research cycle, social network, and research investment ratio/rate on the technology innovation of industrial cluster can also be observed.