Metropolitan Congestion Performance Measures Based on Mass Floating Car Data

Single traffic flow parameters such as speed, volume and density cannot give a better picture of congestion in itself. And broad, general performance measures are always limited by data absence. As the widespread using of GPS devices, floating car data can be collected as a key data source of congestion performance measures with large sample size, wide coverage and high accuracy. It makes it possible congestion performance measures can be done at different road classes and real time. The detail of the floating car data characters and processing procedure are presented. And "five-dimensional" traffic congestion performance measures are produced with congestion intensity, spatial and temporal distribution, congestion occurring frequency (bottleneck detection) and reliability. A scientific approach for quantitative analysis of urban traffic congestion for inner-city and inter-cities is provided. At last, empirical analysis in Beijing and Shanghai of China are deployed, which verify the feasibility and practicality of the measures and the methods.