Analysis of travel-time variation over multiple sections of Hanshin expressway in Japan.

Travel-time variation frequently occurs in urban arterial road networks as a result of demand and supply variations as well as of external factors such as adverse weather and natural disasters. Study of travel-time variation is useful for measuring travel-time reliability studies. The objective of this study was to analyse the factors influencing travel-time variation among multiple sections of the Kobe route of the Hanshin Expressway in Japan. In this study, the Seemingly Unrelated Regression Equation (SURE) model was adopted to analyse the travel-time variation due to traffic volume, traffic accidents and rainfall as factors from the supply side, demand side and external factors of the transportation system respectively. This study identified that there was contemporaneous error correlation among multiple sections of the Kobe route. This error causes significant difference in estimated model parameters between the SURE model and Ordinary Least Square (OLS) model. Particularly, it has been observed that SURE model coefficients obtained for the rainfall parameter were 42% lower than OLS model coefficients. The results of this study emphasize that the SURE model is good for analysing the influence of various factors on travel-time variation among multiple sections of the Hanshin Expressway. [ABSTRACT FROM AUTHOR]