【摘要】 针对现有行人仿真模型在行为特性表示不全面和建筑物空间结构表达、空间单元状态的动态存储及转换、仿真过程及结果可视化等方面存在的缺陷,提出了基于双层协作机制的双向行人流仿真模型.通过对南京地铁新街口站换乘通道内行人行为的实地观测,全面分析了行人在双向通道内的行为特征.在模型中的通道环境层,利用GIS技术进行栅格化并存入空间数据库,在仿真输出阶段应用可视化引擎将场景输出至图形界面;在行人行为层,基于Multi-agent和CA理论建立行人行为决策模型,构建的智能体考虑了行人个体特性和行为.通过计算机仿真,揭示了平均速度与仿真时间步、行人密度的关系.模型成功再现通道内双向行人流的编队成行及相变现象.更多还原

【Abstract】 Pedestrian behavioral characteristics in metro interchange passageways were analyzed by means of the site observation data at Xinjiekou Station of Nanjing Subway.A comprehensive bi-layer cooperative model for bi-directional pedestrian flows was developed to eliminate the drawbacks in the existing models in terms of behavioral characteristics,the spatial structure expression,dynamic storage and transformation of spatial unit states,and the simulation visualization.In the model,the layer of passageways was first rasterized using GIS technology and then the data were stored in the spatial database and finally output to the GUI by visualization engine in the post-simulation stage scene.In the layer of pedestrian behaviors,a decision model was established on the basis of the Multi-agent and CA theories.The model defined the agent with the consideration of pedestrian’s characteristics and actions.The simulation results obtained from the model exercise illustrated the density-speed curve and the time interval and speed curve.The simulation shows that the model can reproduce the typical phenomena observed in reality such as formation of bi-directional pedestrian flows and the phase transition.更多还原