چکیده:
With the increase in traffic on two-lane roads to crowded and dense states or near them, the conversion of two-lane roads to four-lane roads is raised. Although the conversion of two-lane roads to four-lane roads is usually based on the need to provide a better level of service for increased traffic volume, this conversion, in addition to providing a better level of service to users, leads to a reduction in accidents and consequently achieves safety benefits. Therefore, this research deals with analyzing the safety effect by assessing the conversion of two-lane roads to four-lane roads. The aim of this research is to provide an estimate of the safety improvement in the conversion of separate two-lane roads to four-lane roads using the cross-section method. In this research, by modeling accidents on two-lane and four-lane roads, the conversion of two-lane roads to four-lane roads is studied by comparing the results of accident prediction models. To conduct this research, accident data for a 4-year period from 1388 to 1391 for 4 axes of Iranian roads with a total length of more than 500 kilometers, as well as their traffic data, were studied. The results of this research show that by converting two-lane roads to four-lane roads, the accident rate will decrease by more than 60 percent.
خلاصه ماشینی:
Although the conversion of two-lane roads to four-lane roads is usually based on the need to provide a better level of service for transporting increased traffic volume, this conversion, in addition to providing a better level of service to users, leads to a reduction in accidents and consequently results in safety benefits.
Therefore, this research intends to provide an estimate of the safety benefits resulting from the reduction in accidents due to the conversion of two-lane roads to four-lane roads by modeling the conditions of the road sets.
The aim of this research is to provide an estimate of the change in safety effect in the conversion of two-lane roads to divided four-lane roads using the cross-sectional analysis method.
Background and Theoretical Framework of the Research To date, various studies have been conducted to develop road accident prediction models, so that through this way, the factors affecting accidents can be investigated and solutions for improving safety can be provided.
Reviewing the studies of Maher and Summersgill 6 (1996), Karlaftis and Golias (2002), Elvik 7 (2003), and Hauer 8 (2010) in accident modeling shows that for modeling, if overdispersion 10 is absent in the data (response variable), the Poisson model can be used, and otherwise, the negative binomial model should be used.
(2007) "A crash-prediction model for multilane roads", Accident Analysis & Prevention, Vol. 39, No. 4, pp.
(
, 
, 
)
