Real-time urban traffic monitoring with global positioning system-equipped vehicles

Abstract

Real-time traffic conditions are useful information based on which many adaptive traffic solutions work. In this study, the authors present a new approach for real-timely monitoring urban traffic with global positioning system (GPS)-equipped vehicles, which provides estimation of urban traffic conditions in real time. The approach first real-timely collects GPS trace data from GPS-equipped vehicles on the urban road network. Then, it periodically clusters the collected data of several minutes, calculates estimated space mean speed (eSMS) and translates eSMS to smooth indexes (denoting traffic conditions). Compared with existing work, the presented one: (i) applies an effective map matching method to cluster GPS trace data; (ii) excludes traffic signal's misleading influences on traffic condition estimation and (iii) judges traffic conditions based on an estimated critical traffic flow characteristic. Some experiments based on GPS taxi scheduling data of Shanghai, China are provided to demonstrate performance of this work.