An Exploratory Analysis of PAS Characteristics in Solving the Static Deterministic User Equilibrium Traffic Assignment Problem on a Large Scale Urban Network

The standard static deterministic user equilibrium (UE) traffic assignment has long been one of the most intensively applied tools utilized by transportation practitioners particularly for travel forecasting. As is, however, well known, the standard UE formulation provides a unique solution for total link flows but not for route flows. In 2010, Bar-Gera proposed a new algorithm called Traffic Assignment by Paired Alternative Segments (TAPAS) to solve the static deterministic UE traffic assignment in a computationally efficient manner while the first time addressing the issue of route flow uniqueness by incorporating a condition of proportionality. Although TAPAS has been successfully implemented on various scales of transportation networks, characteristics of pairs of alternative segments (PASs) that comprise the solutions of the standard UE model, especially for a large-scale urban network, have not been adequately explored and revealed. This study examines and compares the underlying characteristics of PAS solutions for the Chicago regional network at both aggregate and disaggregate levels for three congestion scenarios applied to a single-class model. Investigations utilize descriptive analysis as well as graphical and map representations of selected PASs to substantiate observations and conclusions. Findings of this study are expected to help guide transportation practitioners in understanding the properties of traffic assignments with unique route flows and in assessing the performance of algorithms in traffic assignment problems