Fitting Time Headway–Vehicle Speed Bivariate Distributions

Abstract

Knowledge of traffic flow characteristics at cross sections helps analysts, software developers, and practitioners to identify, model, and simulate traffic flows as inputs in operational analysis. This work presents a complete procedure for constructing bivariate time headway (TH)–vehicle speed (VS) distributions starting from raw traffic data on two-way two-lane roads. The general framework of the procedure, together with functional components and their mutual interactions, was designed and implemented. The effect of flow rate in both directions (analyzed and opposite) on TH-VS distributions was examined. The possibility of associating the TH-VS probability density function with several kinds of traffic conditions in both directions and lanes is a significant and interesting advance with respect to previous works. The procedure was applied to a set of four cross sections belonging to the road network of the province of Venice (northeast Italy). This application enabled testing of the various steps of the procedure and acquisition of better knowledge concerning the way the findings can be used for traffic flow description and simulation.