Identification of factors related to the severity of highway accidents using binary logistic regression model

In 1399, 584 deaths due to vehicle accidents were recorded in the metropolis of Tehran. Of these, 67% of casualties occurred on highways and 33% on other urban thoroughfares. Therefore, researchers have paid much attention to the analysis and investigation of the causes and severity of accidents in order to help reduce these damages by providing appropriate solutions. The main purpose of this study is to determine the factors related to the severity of accidents that have occurred on highways in Tehran.

In this study, using 90852 real data of ten highway accidents in Tehran in the period of 1390 to 1399, using binary logistic regression method, the factors related to the severity of accidents on Tehran highways were identified.

The results showed that the probability of injury accidents with 57.8% from 12 noon to 4 pm is higher than other time periods. Accidents on weekdays are three times more common than weekend accidents. When other factors were kept to a moderate level, motorcycle crashes had a 79.7% chance of injury. It turned out that the role of speed in accidents is more than 13 times greater than the mechanical defects of the vehicle.

Urban management can now take a number of measures to improve traffic safety, including slowing down and encouraging the use of helmets for motorcyclists and making informed decisions about creating speed zones on highways.