A Model of Saturation Flow Using Traffic Subgroups

Abstract

This thesis presents a methodology for estimating saturation flow rates at signalized intersections by traffic subgroups. A subgroup is defined as a group of vehicles of a specific vehicle classification that make a single directional movement from one lane. The subgroup method is founded on the procedures described in the 1997 Highway Capacity Manual (HCM) (Transportation Research Board, 1997) but extends beyond the HCM's lane group model to provide results that can be aggregated at multiple levels. Rather than assuming homogeneous conditions within each lane or lane group as is the case with many capacity guides, the subgroup method decomposes a traffic stream into individual components, each of which have unique saturation headways. Comparisons with the HCM show that under similar assumptions, the subgroup method produces similar saturation flow rates when aggregated at the lane group level. This gives confidence for applying the subgroup approach to estimate individual lane performance.The most critical element of the subgroup model is the estimation of lane volumes. Lane volume field data were gathered at four sites. Results of an evaluation of lane distribution strategies for estimating lane volumes when a choice is present indicate that the equal back of queue strategy best reflects driver behavior. Based on the observed field data, the equal back of queue strategy outperforms the equal delay strategy and the equal flow ratio strategy which are widely used in various international capacity guides.