交通运输系统工程与信息 ›› 2016, Vol. 16 ›› Issue (5): 192-197.

• 系统工程理论与方法 • 上一篇    下一篇

考虑行人随机行为波动的改进社会力模型

曲昭伟1,曹宁博1,陈永恒*1,白乔文1,赵利英1,康萌2   

  1. 1. 吉林大学交通学院,长春130022;2. 吉林省吉林市公安局交警支队,吉林吉林132000
  • 收稿日期:2016-02-26 修回日期:2016-05-16 出版日期:2016-10-25 发布日期:2016-10-25
  • 作者简介:曲昭伟(1962-),男,吉林省长春人,教授,博士.
  • 基金资助:

    国家自然科学基金项目/National Natural Science Foundation of China(51278220,51278520);吉林省科技发展计划重点项目/ Science and Technology Development Project of Jilin Province(20130206093SF).

A Modified Social Force Model Considering Random Behavior Fluctuations

QU Zhao-wei, CAO Ning-bo, CHEN Yong-heng, BAI Qiao-wen, ZHAO Li-ying, KANG Meng   

  1. 1. College of Transportation, Jilin University, Changchun 130022, China; 2. Traffic Police Detachment, Jilin City Public Security Bureau of Jilin Province, Jilin 132000, Jilin, China
  • Received:2016-02-26 Revised:2016-05-16 Online:2016-10-25 Published:2016-10-25

摘要:

为在仿真模型中更全面地反映行人特性,并研究随机行为对双向行人过街的影响,本文基于社会力模型建立了考虑行人随机行为的仿真模型.分析了随机行为对行人过街的影响,从速度和受力两个方面对行人随机行为进行建模,建立两个不同的行人随机行为模型.通过数值模拟,在双向行人比例1:1 的条件下,得到了考虑随机行为条件下行人的仿真轨迹和行人平均过街时间与行人总数的关系;此外在不同行人随机行为各向异性参数λαε 条件下,对比分析了仿真过街时间与经验模型过街时间.结果表明,行人随机行为导致过街时间增加,考虑随机行为后,改进模型二能够得到与实际相似的过街轨迹,当0.5<λαε <1时,改进模型二的行人仿真过街时间与利用经验模型得到的过街时间最为接近.

关键词: 交通工程, 随机行为波动, 社会力模型, 行人, 人行横道

Abstract:

To make pedestrian microscopic simulation model more comprehensive, this paper builds a pedestrian microscopic simulation model to describe pedestrian crossing process considering random behavior based on social force model. In the modified model, the influence of random behavior on pedestrian crossing is studied. Two models are modeled to represent random behavior from speed and force. Based on the simulation results, on condition that the ratio of bidirectional pedestrian is 1:1, the pedestrian simulation trajectories and the relationship between the crossing time and crossing pedestrian number are acquired. Besides, when random behavior anisotropic character λαε takes different values from 0 to 1, the difference between pedestrian simulation crossing time and calculates crossing time based on crossing time empirical model are compared and analyzed. It is found that random behavior leads the crossing time to increase, and the second modified social force model can reproduce the crossing trajectories similar to actual crossing trajectories. when random behavior anisotropic character λαε takes different values from 0.5 to 1, pedestrian simulation crossing time of second modified social force model are closest to calculated crossing time based on crossing time empirical model.

Key words: traffic engineering, random behavior fluctuations, social force model, pedestrian, crosswalk

中图分类号: