考虑视域影响的疏散行为建模及双向行人流仿真  

Evacuation behaviour modelling and simulation of pedestrian counter flow considering influence of visual field

作　　者：何大治 李晓克 李明明 HE Da-zhi;LI Xiao-ke;LI Ming-ming(School of Civil Engineering and Communication,North China of Water Resource and Electric Power,Zhengzhou 450045,China)

机构地区：[1]华北水利水电大学土木与交通学院,河南郑州450045

出　　处：《浙江大学学报（工学版）》2020年第6期1185-1193,共9页Journal of Zhejiang University：Engineering Science

基　　金：国家留学基金委资助项目(201708410400);河南省高等学校重点科研资助项目(16A580004).

摘　　要：通过实验获取行人视域半径数据,根据不同视域半径下的碰撞检测结果,建立包括转向、跟随、减速、停止和正常移动等行人微观行为的决策算法.针对空间连续性导致的人员重叠的问题,设计消除行人位置重叠的算法.设计通道双向行人流仿真实验,仿真过程显示,模型能较好地实现自动渠化现象,并消除行人位置重叠的问题.在行人流密度较高的情况下,对比实际实验与仿真实验的自动渠化过程,发现两者形成的行人通道行数基本一致,疏散时长接近.在行人流密度较低的情况下,仿真实验得到的行人通道行数与计算结果一致.仿真实验的行人流量-密度数据拟合曲线与调研数据拟合曲线吻合度较高.Pedestrian visual radius data was obtained through experiments.A decision algorithm for pedestrian micro behaviors,including turning,following,decelerating,stopping,and normal movement,was established based on collision detection results at different sight radiuses.An overlap elimination algorithm was designed for the problem of personnel overlap caused by spatial continuity.A pedestrian counter flow simulation experiment was designed.The simulation process show that the model can achieve the automatic channelization well;while,the problem of overlapping pedestrian is also eliminated.In the case of high pedestrian flow density,the numbers of pedestrian lanes formed by the simulation and actual experiment are basically the same,and the evacuation durations are close.In the case of low pedestrian flow density,the number of pedestrian lanes formed by the simulation experiment is consistent with the calculation results.The fitting curve of pedestrian flow-density data in the simulation experiment agrees well with the fitting curve of the survey data.

关 键 词：行人疏散 空间连续模型 AGENT建模 视域 行为决策 渠化现象 

分 类 号：U121[交通运输工程]