交通探测车的样本量模型  

Sample Size Model of Traffic Probe Vehicles

作　　者：胡小文[1,2] 杨东援[1] 

机构地区：[1]同济大学交通运输工程学院,上海201804 [2]常州市规划设计院,江苏常州213003

出　　处：《公路交通科技》2009年第10期111-116,共6页Journal of Highway and Transportation Research and Development

基　　金：国家自然科学基金资助项目(50738004)

摘　　要：全面回顾了探测车样本量问题的研究,分析了现有模型中的不足,并从路段和路网两个层面探讨了交通探测车样本量问题。将路段样本量问题归结为一个有限总体的无放回简单随机抽样问题,并给出了修正的路段样本量模型。结果表明,修正的路段样本量模型确定的最小样本量小于现有的模型。引入了路网覆盖率,给出了路网样本量的通用模型,该模型将路段样本量和路网样本量统一起来。以国内某大城市出租车GPS数据为例,建立了在不同的时段特征和路段样本量要求下,路网覆盖率与探测车总数量之间的函数关系。The researches on problem of sample size of probe vehicle were reviewed and some shortcomings of existing models were analyzed. The sample size model of probe vehicles in terms of link and road network was discussed.The link sample size of probe vehicle was regarded as simple random sampling without replacement within finite population, and an improved link sample size model of probe vehicles was proposed. The result shows that the minimum sample size determined by the improved model is no more than that determined by existing models.The concept of road network coverage rate was introduced and a general model of road network sample size was proposed to unify the sample sizes of link and road network. In the end, the function reflecting the relationship between the road network coverage rate and the number of probe vehicles under the condition of different periods and sample sizes of link was established based on a metropolitan GPS probe vehicle data.

关 键 词：智能运输系统 样本量模型 曲线拟合 探测车 路网覆盖率 

分 类 号：U491.116[交通运输工程—交通运输规划与管理]