对Repast以及由RepastS构建仿真模型的基本流程作了简单介绍，并以昆明市主城区为例，以RepastS松散集成GIS，构建了一个反映城市人口空间分布的仿真模型。模型首先将昆明市主城区的324万总人口分为3240个Agent，选取对人口空间分布影响较大的交通、经济、行政驻地、企事业单位、水体、坡度六类因素作为影响因子，运用主成份分析功能以及专家打分的方式对各影响因子分别赋予对人口空间分布的影响权重以及自身的影响度值，建立影响Agent随机移动以及停留的函数模型。模型根据Agent最终停留的坐标自动计算落在每一个行政区内的Agent个数。将模拟结果与昆明市主城区实际人口分布数据进行对比，结果表明模型的仿真度较高，证明模型的构建是成功的。%This article briefly introduced Repast and the basic method and process to simulate simulation model based on Repast S. And taking the main urban area of Kunming City as an example, a simulation model was constructed to reflect city's population distribution. In this model, GIS and Repast S were integrated loosely, 3240000 of the total population of the main urban area of Kunming City was divided into 3240Agent, each Agent according to certain rules of conduct random moving in the six kinds of influence factor: economy, traffic, en- terprises and institutions, rivers, Slope. Each influence factor were given weight values according to the influence on the spatial distribution of population and itself by the methods of the principal component analysis and expert e- valuation. On the basis of the weight values, the agent random movement and residence function is construc- ted. Model according to the Agent final stay coordinate automatic count the Agent number of final fall in each ad- ministrative district. The simulation results were compared with Kunming city actual population distribution data, the result shows that this model has high accuracy, proved that the model construction is successful, the model is feasible and available.
展开▼
