Loading...

Effects of Vehicle Characteristics on Accident Severity; Comparison Between Different Types of Accidents

Najaf, Pooya | 2012

426 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 42806 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Kermanshah, Mohammad
  7. Abstract:
  8. Evaluation of the effects of vehicle characteristics on accident severity is the main objective of this research. In addition to some static and dynamic features of passenger cars, some characteristics of driver and roadway are studied, too. In this research, driver injury severity is modeled as accident severity index. First, the severity of head-on, rear-end, angle and sideswipe accidents are modeled using Binary Logit models. Then, Multinomial Logit models are applied to develop driver injury severity models for head-on and rollover accidents. Finally, Nested Logit model is developed to model severity of head-on accidents to study the performance of different types of Logit models. Furthermore, Artificial Neural Networks model is trained to predict severity of head-on accident to compare their prediction ability with Logit models. According to the results, increase in vehicles’ weight, length, height and wheelbase, and decrease in vehicles’ width will generally intensify the severity of different types of accidents. Also, vehicle weight is one of the most important factors in determination of accident severity. In addition, seatbelt and airbag are effective in decreasing driver injury severity, and older drivers are injured more probably. Based on the final results, vehicle mismatch is the most important cause of severe injuries in most types of accidents. The modeling results show that Nested Logit is the most appropriate structure to model driver injury severity, and Artificial Neural Networks do not have any priority to discrete choice models in accident severity prediction and analysis.
  9. Keywords:
  10. Crash Severity ; Artificial Neural Network ; Choice Model ; Vehicle Characteristics

 Digital Object List

 Bookmark

No TOC