In order to improve the traffic safety in highway tunnels, thirteen factors of environment, drivers, roads and vehicles are extracted as independent variables based on real-world crash reports. And the severity (“ casualties ” or “ no casualties ”) and configuration (“ a multi vehicle crash ” or “ a single vehicle crash ” ) of crashes are taken as dependent variables. In order to study the correlation of two dependent variables and identify significant factors affecting the crash severity and configuration, a bivariate probit model considering the internal correlation of dependent variables is constructed. The results indicate that there is a positive correlation between the configuration and severity of traffic crashes in highway tunnels: the fatality risk of multi-vehicle crashes is higher. Besides, the higher the proportion of trucks and the higher the grade of highways, the severer the crashes. The more heavy trucks, the more likely it is to cause single vehicle crashes. The results of the current study could provide some valuable insights for improving traffic safety in highway tunnels.

CICTP