Assessment of critical headway estimation methods across multiple unsignalized intersections

Abstract

Driver behavior under local conditions has a significant influence on the value of accepted headways required for performing minor maneuvers. In operational analyses and the design of unsignalized intersections, driver behavior is quantified through the value of the critical headway. It is assumed that the critical headway represents the shortest headway that a driver is willing to accept. Since it is not possible to measure the critical headway directly from field observations, measurement results are processed using application software and mathematical tools. To determine the critical headway at three priority intersections in the urban area of the city of Doboj, six different methods were applied: Harders’, Logit, Lag, Raff’s, Wu’s, and the Maximum Likelihood method. In order to identify which of the applied methods provides the best solution for the analyzed intersections, a comparative analysis was conducted between the theoretically obtained capacity values and the actual measured capacity value. The Lag method was the most suitable for intersection I1, the Maximum Likelihood method was chosen for intersection I2, while the Raff method was selected for intersection I3.