摘要
【目标】为精细化描述行人在信号控制路段人行横道处的过街行为,分析研究行人过街的行为机理,基于决策层级理论构建行人过街行为模型。【方法】在战术层级上,考虑到行人的过街过程可划分为路侧至路中、路中至路侧两个阶段,基于二元Logistic回归模型对行人在路侧与在路中不同场景下的过街决策分别建模。在执行层级上,考虑传统社会力模型在模拟信号控制条件下行人过街运动的局限性,通过对不同信号阶段的期望速度差异化标定、引入对向避让力来改进社会力模型,进而构建行人过街微观运动模型。最后,利用AnyLogic仿真实现过街行为模型,结合实际观测数据,验证了所建模型的有效性。【结果】路侧环境下和路中环境下行人过街决策的影响因素差别较小,但是行人在路侧至路中阶段所做决策考虑的因素要更加全面,还包括道路中央分隔形式和安全岛设置两个因素;行人性别、同向过街行人数量、车流量、车辆到达时间、过街距离、过街绿灯时长占比等因素对行人过街决策结果的影响较为显著;仿真得到的与实际调查的相应信号阶段行人过街的平均速度均服从同一分布。【结论】基于决策层级理论构建的行人过街行为模型更为细致地反映行人过街时的决策行为,描述不同信号状态下行人期望速度的差异;相比传统模型,引入对向避让力能更真实地模拟行人过街过程中的微观运动行为。
[Objective]To refine the description of pedestrian crossing behavior at signalized midblock crosswalks,and analyze the mechanism of pedestrian crossing behavior,a pedestrian crossing behavior model was constructed based on decision hierarchy theory.[Method]At the tactical hierarchy,a pedestrian crossing process was divided into two phases,i.e.,from roadside to median strip,and from median strip to roadside.The pedestrian crossing decisions were modeled respectively on roadside and median strip based on the binary Logistic regression model.At the execution hierarchy,due to the imperfections of classical social force model for pedestrian motions simulation in signalized conditions,the social force model was improved through the differential calibration of expected speed at different signal phases,and the introduction of opposing avoidance force,thereby constructing the microscopic pedestrian crossing motion model.Finally,the crossing behavior model was realized through AnyLogic simulation.The model effectiveness was verified with the actual observation data.[Result]The influencing factors of pedestrian crossing decisions in both roadside and median strip scenarios are less different.The factors,considered by pedestrians in making decisions from roadside to median strip,are more comprehensive,i.e.,road central separation,and refuge island.The factors,having significant influences on pedestrian crossing decision,are gender,number of crossing pedestrians in same direction,traffic flow,vehicle arrival time,crossing distance,and proportion of green light duration.The average speed of pedestrians crossing during the corresponding signal phases obtained from simulation and actual investigation both follow the same distribution.[Conclusion]The decision hierarchy theory based pedestrian crossing behavior model more finely reflects the decision-making behavior of pedestrians crossing.It can describe the difference of pedestrian expected speeds at different signal phases compared with traditional models,the opposing avoidance force would more realistically represent the micro motion behavior of pedestrians during crossing.
作者
闫伟曦
陈鹏
唐鹏
刘霆威
YAN Weixi;CHEN Peng;TANG Peng;LIU Tingwei(School of Transportation and Logistics Engineering,Wuhan University of Technology,Wuhan,Hubei 430063,China;CCCC Second Highway Consultants Co.,Ltd.,Wuhan,Hubei 430056,China)
出处
《公路交通科技》
北大核心
2025年第11期11-20,共10页
Journal of Highway and Transportation Research and Development
基金
国家自然科学基金项目(51208400)。
关键词
智能交通
行人过街仿真
社会力模型
行人过街行为
决策层级
intelligent transport
pedestrian crossing simulation
social force model
pedestrian crossing behavior
decision hierarchy
