Study on Space-Time Distribution Characteristics of Floating Car Data Based on Large Samples

Abstract

Floating Car Data has been used to evaluate traffic conditions in Real-time Transportation Information Systems. In these systems, taxis are often used as probe cars. Because of the randomicity of taxis when travelling in cities, it is necessary to analyze taxis’ distribution characteristics in road networks. In this paper, two indexes named Detecting Intensity and Detecting Rate are designed to analyze the Space-Time Distribution Characteristics. 5,000 taxis in Hangzhou of China are selected as probe cars, and Floating Car Data are continuously collected from these taxis during a week. The conclusions show that Detecting Intensity and Detecting Rate can clearly demonstrate the Space-Time Distribution Characteristics in different time and road types. Urban express ways, arterial roads can usually be detected by probe cars with more dependability. Traffic conditions evaluated through FCD in peak hours on a day is probably more dependable than in off-peak hours. At the same time, the relationship between distribution characteristics and sample size are also analyzed, in order to help find a more reasonable probe car sample size for Real-time Transportation Information Systems.