A Study of Structure of Road Accidents Status Prediction Model for Highways

Abstract—With regard to the importance of road accidents and the resulting costs, which include human, economic and social losses, the study and modeling of road accidents will be inevitable in identifying the causal factors of accidents. This article analyzes the construction of a model for the severity of fatal road accidents taking into account of the characteristics of the driver, motor vehicle, road and the environmental conditions of the roads of the country. In order to construct the accidents severity model, the Logit function of fatality likelihood is used. The function’s output varies between zero and one. As the output becomes closer to one the likelihood of fatality will be more, and as it approaches the zero the likelihood of fatality will come down. In such models, such factors as elderly drivers, the existence of motorcycles and bicycles, inattention to the front, exceeding the authorized speed, unauthorized deviation to left due to overtaking, passing unauthorized place, illiteracy of drivers, etc. will increase the severity of accidents, and such factors as clear weather, standard road width, etc. will decrease fatal accidents. The factors are discussed in detail in this article.