Traffic State Identification Considering Differences between Road Segments and Intersections

Abstract

Present studies on urban road traffic state identification usually set road segments and connected intersections as an entirety. However, traffic flow on road segments and intersections has significantly different characteristics, so it may cause identification error when implementing the same identification criterion. This paper divides road segments and intersections respectively into small units for the purpose of traffic state differential identification. Then, statistical analysis is used to compare the different characteristics of traffic flow between road segments and intersections, during the process, real floating car data of Nanjing is used. The outcome clearly shows the differences and proves that differential identification is needed. Also it contributes to the making of different identification criterion: median speed and ratio of maximum queue length to unit’s length are used as the indexes of state identification respectively for road segments and intersections. A case study is conducted to prove the feasibility of the identification method, and the conclusion is that differential traffic state identification can better reflect a real traffic flow state and largely reduce identification errors.