Research On Travel Time Estimation At Signalized Segment Based On Small Sample Floating Car Data

An Effective transportation system is important for development of modern city. From 70's of last century, transportation problems such as severe congestion and low efficiency, which are followed by a series of environment and social problems, restrict the development speed of modern citys to some extent. In view of this point, the concept of Intelligent Transportation System (ITS) is introduced to the whole world. Traffic detection technique, as a fundamental component of many applications in ITS, has important impact on the operation efficiency of ITS applications.This thesis introduces traffic detection technique, especially on travel time collection methods existed nowadays firstly. Then different travel time estimation methods including estimated from traffic detectors and from float car data are overviewed in detail. By pointing out problems in currents methods, a travel estimation method based floating car data of small sample size is introduced in this thesis. This method can reduce errors caused by data fluctuation induced by signal period in short arterial road in center area of big citys. Finally, based on real project, empirical analyses are provided to prove the efficiency of this method. Analysis result shows the consistency of travel time estimated from detectors and float car data. Moreover, comparison between errors in estimated travel time from this method and common method gives another proof for method efficiency. Conclusions are drawn that the method provided in this thesis can improve travel time estimation accuracy.