【摘要】 应用车载GPS探测车的运行数据来估计道路交通状态是目前交通检测领域的新技术。从该技术基本概念出发,重点对其中的数据采集和交通参数估计问题进行了探讨。结合英国浮动车数据系统(FVD)数据采集的优化模式及目前探测车最小覆盖率的研究,讨论了数据采集的效率与样本大小的问题。分析了目前典型的GPS探测车交通参数估计模型,如多元回归、神经网络、模糊推理及速度积分等模型。对该技术的进一步发展及在我国的应用前景进行了分析和展望。更多还原

【Abstract】 For GPS equipped probe vehicles to be a distinct source of real-time traffic information,it is a new solution to assess real-time traffic performance using data from them.This paper introduces its basic principle.The issues about the efficiency of data collection and sample size are addressed,through an example of the optimum data collection model of Floating Vehicle Data(FVD) in Britain and a discussion of the issues involved.Typical approaches to estimate traffic parameter in this solution are analyzed,such as multiple regression analysis,neural network,fuzzy inference and integral of speed data.Finally,the key issues of its application in China are prospected.更多还原