基于HErD分布的信号交叉口车辆到达规律动态预测  

Dynamic Prediction of Vehicle Arrival Law at Signalized Intersection Based on HErD Distribution

作　　者：蒋阳升[1,2] 韩世凡[1,2] 张改 胡路[1,2] 

机构地区：[1]西南交通大学交通运输与物流学院,四川成都610031 [2]交通运输智能化国家地方联合工程实验室,四川成都610031 [3]江苏智通交通科技有限公司,江苏南京210000

出　　处：《石家庄铁道大学学报（自然科学版）》2015年第3期105-110,共6页Journal of Shijiazhuang Tiedao University(Natural Science Edition)

基　　金：国家自然科学青年基金(51108391)

摘　　要：为掌握信号交叉口实时优化信号控制和交通诱导所需的车流到达信号交叉口的随机时变规律，利用HErD分布具有无限逼近任意非负随机变量的特性，提出基于HErD分布的信号交叉口车辆到达时间间隔变化规律动态预测的方法，并以一个交叉口的2个进口道进行实例验证分析。实验结果为：其中方向一的拟合度范围为86．17％～93．51％，平均拟合度高达90．32％，方向二的拟合度范围为93．24％～98．52％，平均拟合度高达96．93％，其高精度的拟合结果表明基于HErD分布的信号交叉口车辆到达规律动态预测具有科学性和实用性。To grasp the random time-varying regularity of arrival traffic at signal intersection,which is needed to control the real-time optimization of signalized intersection and traffic guidance,with the unlimited approximate non-negative random variable characteristics of the HEr D distribution,this paper proposes a dynamic prediction method of vehicle arrival time distribution changes based on the HEr D distribution. And an intersection of two import channel is instantiated to verify and analyze. Experimental results: Direction one fits a range of86. 17% ~ 93. 51%,with an average of up to 90. 32 percent fit; Direction two fits a range of 93. 24% ~ 98.52%,with an average of up to 96. 93%. The precision fitting results show that the dynamic prediction method of vehicle arrival time distribution changes based on the HEr D distribution is scientific and practical.

关 键 词：信号交叉口 车辆到达规律 混和爱尔朗HErD分布 动态预测 

分 类 号：U491.4[交通运输工程—交通运输规划与管理]