TY  - JOUR
PY  - 2017//
TI  - Development of running red light crash modelling at signalized intersections for safety evaluation
JO  - Journal of transport research (Seoul)
SP  - 41
EP  - 51
VL  - 24
IS  - 2
N2  - According to the 2015 statistics of the National Police Agency, running red light (RLR) crashes represent a serious safety problem at signalized intersections throughout South Korea. The purpose of this study is to comparatively analyze the correlation between RLR crashes and various drivers, vehicles, and road environmental factors for a comprehensive safety evaluation. This study researched factors using a binary logistics regression analysis with SPSS 20.0 software. Data was used from 4,221 RLR crashes which occurred at 73 intersections in the cities of Suwon, Ansan and Sungnam between 2012 and 2014. The main results showed that 81.5% of RLR crashes were caused by male drivers, after sun down, while traveling straight. Second, the RLR crashes' killed or seriously injured were clearly defined by main street traffic volume, intersection area, yellow time gap, and nighttime average vehicle speed. Finally, a statistically significant model was developed using the Hosmer and Lemeshow test giving an output of 6.62.<p />  <p>Language: ko</p>
LA  - ko
SN  - 1738-4028
UR  - http://dx.doi.org/
ID  - ref1
ER  -