Prediction of passing probability and risk evaluation along undivided road under heterogeneous traffic

Abstract

This study focuses on the safety part of the passing manoeuvre. Passing is considered as one of the most complex, unavoidable and challenging manoeuvres on two lane roads. A binary logistic regression model was developed to determine the probability of passing or aborting based on field data. The result shows that the probability of passing is affected by the factors such as speed of passing vehicle, speed of overtaken vehicle, speed of opposing vehicle, clearance time, and the total passing time. Risk assessment is an important aspect which should be addressed while assessing the passing operations. In the present study, a risk chart was developed based on clearance time and vehicle interaction. An overall risk table was developed for all possible combinations of risk. The study revealed that the selected road stretch along NH-66 in Kerala, India is more susceptible to passing crashes due to insufficient clearance time. This paper brings out a rational methodology for calculating a safety measure in terms of risk index factor which could be used in simulation softwares for adequate passing sight distance operational analysis.