Loading...

Estimation of the Number of Accidents using TCT Traffic Conflict Technique and EVT Extreme Value Theory

Mirza Ardestani, Hamed | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 56811 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Nassiri, Habbibollah
  7. Abstract:
  8. The first step in improving traffic safety and reducing damage to road users is to evaluate traffic safety and identify black spots, which is possible in statistics-based methods using traffic accident data from previous years. One of the disadvantages of statistical methods is the existence of a series of limitations in collecting accident data and spending a lot of time and money. The purpose of this research is to provide a technique to evaluate traffic safety and estimate the number of accidents using the traffic conflict technique and extreme value theory modeling. The method of conducting this research is based on image processing, recording, and analysis of the conflict index of two road users at the entrance of Shkouhieh Industrial Estate of Qom. At first, field images were prepared from the desired location and imported to T-Analyst software for analysis. Then by determining the type (for example, rear-end and sidewipe), the number and severity of conflicts, and indicators such as time to collision TA and conflicting speed CS, a preliminary qualitative analysis was carried out and road safety problems were identified to provide safety countermeasures. In the next step, using extreme value theory, two models were built with the approaches of BM blocks and pick-over threshold POT for 4 conflict indicators such as PET, TTC, T2, and Tadv to estimate the number of annual accidents. Finally, the actual number of annual accidents was compared with the output of the models to determine the relationship between the conflict indicators with the number of annual accidents and the best indicator for estimating the number of annual accidents
  9. Keywords:
  10. Traffic Safety ; Traffic Accidents ; Extreme Value Theory ; Block Maxima ; Peak Over Threshold ; Traffic Conflict

 Digital Object List

 Bookmark

No TOC