基于轨迹数据的无信号交叉口行人过街行为研究
|
摘要
随着我国城市道路运行效率的降低和交通安全形势的日益严峻,行人交通安全已成为普遍关注的问题。本文从无信号道路交叉口的实测轨迹数据中提取相关参数,建立基于随机森林算法的行人过街决策模型并进行效果检验和验证,模型正确率可达93.1%。同时挖掘出各个解释变量对过街决策的重要度,证明将随机森林算法用于行人过街决策分析具有较好的拟合度与预测精度。本研究可应用于人车交互仿真工具中模拟行人行为,为进一步研究人车混行的无信号道路交叉口仿真和控制系统奠定了基础。
The low operating efficiency and severe situation of urban traffic lead to pedestrians' safety problems. Based on observed trajectory data at an unsignalized intersection, this work analyzed pedestrians' crossing behavior. The decision-making behavior of pedestrians was modeled using random forest algorithm, and the model was then calibrated and validated according to empirical data, which accuracy reached 93.1%. This work has found several key factors on the decision-making behavior of pedestrians, which were never reported before, verified that random forest model had a great goodness of fit and prediction. The research can be applied to simulation tools, which laid the foundation of further study on the simulation of unsignalized intersections.
The low operating efficiency and severe situation of urban traffic lead to pedestrians' safety problems. Based on observed trajectory data at an unsignalized intersection, this work analyzed pedestrians' crossing behavior. The decision-making behavior of pedestrians was modeled using random forest algorithm, and the model was then calibrated and validated according to empirical data, which accuracy reached 93.1%. This work has found several key factors on the decision-making behavior of pedestrians, which were never reported before, verified that random forest model had a great goodness of fit and prediction. The research can be applied to simulation tools, which laid the foundation of further study on the simulation of unsignalized intersections.
引文
[1]廖明军,王凯英,姜卉.无信号控制行人过街行为观察和反应时间分析[J].综合运输,2014(4):59-62.
[2]Yang J,Deng W,Wang J,et al.Modeling pedestrians'road crossing behavior in traffic system micro-simulation in China[J].Transportation Research Part A,2006,40(3):280-290.
[3]Lavalette B C D,Tijus C,Poitrenaud S,et al.Pedestrian crossing decision-making:a situational and behavioral approach.[J].Safety Science,2009,47(9):1248-1253.
[4]Khatoon M,Tiwari G,Chatterjee N.Impact of grade separator on pedestrian risk taking behavior[J].Accident Analysis and Prevention,2013,50(3):861-870.
[5]邢胜.非均衡学习及在产前筛查高风险预测中的应用[D].保定:河北大学,2017.
[6]Phua C,Alahakoon D,Lee V.Minority report in fraud detection:classification of skewed data[J].Acm Sigkdd Explorations Newsletter,2004,6(1):50-59.
[7]Breiman L.Random Forests[J].Machine Learning,2001,45(1):5-32.
[8]高珍,柯阿香,余荣杰,等.基于随机生存森林的交通事件持续时间预测[J].同济大学学报(自然科学版),2017(9):1304-1310.
[9]沈涛.基于视频检测的行人交通参数提取技术研究[D].北京交通大学,2011.
[10]明均仁,肖凯.基于R语言的面向需水预测的随机森林方法[J].统计与决策,2012(9):81-83.
[11]孙智勇,荣建,何民,等.信号交叉口人行横道处的行人可接受间隙研究[J].公路交通科技,2004,21(11):102-104.
[12]赵伟涛,钱大琳,杨彦青.行人穿越人行横道临界间隙研究[J].武汉理工大学学报(交通科学与工程版),2012,36(4):808-811.
[13]朱芳芳.无信号控制交叉口行人过街间隙选择模型研究[J].交通信息与安全,2011,29(2):36-39.
[2]Yang J,Deng W,Wang J,et al.Modeling pedestrians'road crossing behavior in traffic system micro-simulation in China[J].Transportation Research Part A,2006,40(3):280-290.
[3]Lavalette B C D,Tijus C,Poitrenaud S,et al.Pedestrian crossing decision-making:a situational and behavioral approach.[J].Safety Science,2009,47(9):1248-1253.
[4]Khatoon M,Tiwari G,Chatterjee N.Impact of grade separator on pedestrian risk taking behavior[J].Accident Analysis and Prevention,2013,50(3):861-870.
[5]邢胜.非均衡学习及在产前筛查高风险预测中的应用[D].保定:河北大学,2017.
[6]Phua C,Alahakoon D,Lee V.Minority report in fraud detection:classification of skewed data[J].Acm Sigkdd Explorations Newsletter,2004,6(1):50-59.
[7]Breiman L.Random Forests[J].Machine Learning,2001,45(1):5-32.
[8]高珍,柯阿香,余荣杰,等.基于随机生存森林的交通事件持续时间预测[J].同济大学学报(自然科学版),2017(9):1304-1310.
[9]沈涛.基于视频检测的行人交通参数提取技术研究[D].北京交通大学,2011.
[10]明均仁,肖凯.基于R语言的面向需水预测的随机森林方法[J].统计与决策,2012(9):81-83.
[11]孙智勇,荣建,何民,等.信号交叉口人行横道处的行人可接受间隙研究[J].公路交通科技,2004,21(11):102-104.
[12]赵伟涛,钱大琳,杨彦青.行人穿越人行横道临界间隙研究[J].武汉理工大学学报(交通科学与工程版),2012,36(4):808-811.
[13]朱芳芳.无信号控制交叉口行人过街间隙选择模型研究[J].交通信息与安全,2011,29(2):36-39.