Title Establishment and Application of Safety Performance Functions for Sections of Highway Tunnels
Year 2018
Degree Master
School Department of Transportation and Logistics Management National Chiao Tung University
Author
Summary The number and frequency of traffic crashes on roads are the common indices for evaluating traffic safety of the road. These indices utilize multi years of accident data for traffic crash analysis. However, most analyses focus on general road sections. Only few had mentioned the difference between entering the tunnel or leaving the tunnel as compared to inside the tunnel. This study focuses on the crash attributes of sections of tunnels, by separating them into normal and continuous tunnel. Normal tunnels have three sections including before, within and after sections. Continuous tunnels have four sections including before, within, after and between sections. This study first calculated the CMF of different sections. Secondly, following the method suggested by HSM, negative binomial distribution was used to establish the SPF considering AADT, segment length, section location and speed limit. The index of effectiveness by estimated SPF was then plotted to show the different effectiveness of the sections. It is found that all the sections aside from within section have CMF smaller than 1. The SPF’s of normal and continuous tunnels were established. For normal tunnels, the index of effectiveness shows that when the speed limit is smaller than 90 km/h, the observed frequency of crashes is mostly greater than predicted crash frequency in every section. However, when the speed limit comes to 90 km/h and greater than 90 km/h, the observed frequency of crash are mostly smaller than predicted crash frequency in every section. For continuous tunnel, the index of effectiveness shows that the observed frequency of crashes is mostly close to the predicted crash frequency in every section. Only when the speed limit is under 90 km/h and lower AADT, the observed frequency of crashes is smaller than the predicted crash frequency.
Hashtags
View count: 191