Date of this Version
There is a noticeable difference between different road users, specifically between passenger vehicles and heavy vehicles such as its length and weight. The majority of previous research were focused on general highway traffic that included passenger cars, trucks, buses, motorcycles, etc. Moreover, HRGC safety studies of specific types of vehicles are relatively few and heavy vehicle safety at grade crossing is even more under-explored.
This research thus focuses on the following objectives: Identify factors related to different injury severity levels of heavy-vehicle drivers (truck/truck-trailer) drivers in crashes reported at HRGCs; to identify a more suitable statistical model for injury severity modeling of truck involved crashes. This study considered variables that have not been explored in previous injury severity studies of truck-involved crashes at HRGCs. Three unordered response models: Multinomial Logit model (MNL), Nested Logit model (NL) and Mixed Logit model (RPL) were evaluated to investigate injury severity of drivers of heavy-vehicles involved in crashes at HRGCs.
Based on criteria used for judging the models and the dataset used in this study, it was concluded that the RPL was most suitable for modeling truck drivers’ injuries in crashes reported at HRGCs amongst the models considered. Truck drivers’ injuries in crashes reported at HRGCs are positively associated with speed of train and road user (truck/trailer), truck-train crash, when train strike road user (truck/trailer), hazardous materials by either one or both users, driver behavior “went around the gates”, age of driver, crashes reported in rural areas and crashes at minimum crossing angle of 60-90 degrees. Whereas truck drivers’ injuries are negatively associated with train detection system, gates, if the track is signaled, when the track is obstructed, HRGCs within 500 feet of a highway and position of vehicle “heavy vehicle stopped on the crossing”.
Advisor: Aemal J.Khattak