Measuring Safety for Urban Tunnel Entrance and Exit Based on Nanoscopic Driving Behaviors

Abstract

The entrance and exit zones of urban tunnel have been considered as the most dangerous parts along the tunnel for its sharp changes of the driving environment. The objective of this paper is to extract a comprehensive measure from numerous original measures which reflect drivers' Nanoscopic behaviors to measure the safety of urban tunnel entrance and exit. Field test was conducted at Xi'an men tunnel located in Nanjing. The drivers' heart rate (HR) and eye movement, together with the pupillary diameter at the entrance and exit of urban tunnel were collected synchronously with the D-Lab system. Operating speed was recorded by the video camera, and then the individual vehicle acceleration was calculated. Following the factor analysis procedure, three factors, which explain 90.18% and 89.15% of the variance in the original data for entrance and exit separately, are retained from the initial four Nanoscopic driving behavior measures. According to weight score of each factor, a comprehensive measure (FE) which could reflect Nanoscopic driving behavior was extracted by linear combination of the three retained factors. To measure the safety level of urban tunnel gateway, FE is classified into three levels. The criterion for safety classification is given as: |FE|=0.05, safe, 0.050.10, dangerous. The validation by comparing with tunnel environmental shows that the measure proposed in this paper is acceptable and more accurate in evaluating the safety of tunnel gateway zones.