Effects of highway geometric elements on accident modelling

Abstract

The tremendous traffic growth generally observed in road transportation has led to a lot of negative consequences in the form of road accidents both in developed and developing
countries. This observation calls for considerable attention towards development a system for the road safety mechanism of rural highway. Road accident prediction plays an important role in accessing and improving the road safety. Fuzzy logic is one of the popular techniques in
the broad field of artificial intelligence and ability to improve performance similar to human reasoning and describe complex systems in linguistic terms instead of numerical values. In this thesis, a system was established based on Fuzzy Inference System (FIS) in which output data such as traffic Accident Rate (AR) and input data such as various highway geometric elements. The study was conducted on two road segments from plain & rolling terrain highway and two road segments from mountainous & steep terrain highway within the rural area of the Indian Territory. Two Highway Accident Rate Prediction Models (HARPMPRT and HARPMMST) were developed due to the complexity of geometric elements of rural highway on different terrain conditions which take horizontal radius, superelevation, K-value, vertical gradient and visibility as input variables and Accident Rate (AR) as output variables. The findings show that the proposed model can be effectively applied as a useful Road Safety tool capable of identifying risk factors related to the characteristics of the road and great support to the decision making of incident management in Intelligent Transportation Systems. Significant positive relationships were also identified between the geometric elements and accident rate. A simulation study and real life data analysis are performed to demonstratemodel fitting performances of the proposed model.