车头时距分布函数的验证、分析与选择

发布时间：2018-04-14 05:16

本文选题：交通工程 + 车头时距　；参考：《公路交通科技》2014年05期

【摘要】：采用30组高速公路实测数据,对常见的车头时距分布函数进行验证。为明确其适用性,分析了分布函数的特征,并提出选择策略。研究结果表明负指数分布能较好地反映车辆到达的随机性,且较为简单,当车道交通量小于250 veh/h时优先选用。移位负指数分布通过最小车头时距的合理取值减小了拟合误差,而爱尔朗分布的阶数必须取整数,扩大了误差,当交通量位于250~750 veh/h时,应优先选用前者。M3分布主要用于拟合交通量大于750 veh/h且车队现象比较明显的交通流,根据线性回归得到的自由车比例α与交通量的函数关系,规定α应小于0.9。改进的M3分布适用性很广,但参数取值比较困难,可作为其他分布函数的有益补充。
[Abstract]:Thirty groups of highway data were used to verify the common time-distance distribution function.In order to clarify its applicability, the characteristics of distribution function are analyzed and the selection strategy is put forward.The results show that the negative exponential distribution can well reflect the randomness of vehicle arrival, and is relatively simple. When the traffic volume is less than 250 veh/h, it is preferred.The shift negative exponential distribution reduces the fitting error by the reasonable value of the minimum headspace, while the order of the Allan distribution must be integer to enlarge the error. When the traffic volume is at 250 ~ 750 veh/h,The former. M3 distribution should be used to fit the traffic flow which the traffic volume is more than 750 veh/h and the phenomenon of the motorcade is obvious. According to the function relation between the free car ratio 伪 and the traffic volume obtained by linear regression, it is stipulated that 伪 should be less than 0.9.The modified M3 distribution is widely applicable, but it is difficult to select the parameters, which can be used as a useful supplement to other distribution functions.
【作者单位】：东南大学交通学院;华中科技大学土木工程与力学学院;
【基金】：国家自然科学基金项目(51208222) 江苏省交通科学研究计划项目(2011-Y-30)
【分类号】：U491.1

【参考文献】