TRANSIMS Implementation for a Small Network and Comparison with Enhanced Four-Step Model

Authors

  • Mansoureh Jeihani Morgan State University
  • Anam Ardeshiri Morgan State University

DOI:

https://doi.org/10.5399/osu/jtrf.53.1.4216

Abstract

Travel demand forecasting is a major tool to assist decision makers in transportation planning. While the conventional four-step trip-based approach is the dominant method to perform travel demand analysis, behavioral advances have been made in the past decade. This paper proposes and applies an enhancemnt to the four-step travel demand analysis model called Sub-TAZ. Furthermore, as an initial step toward activity-based models, a TRANSIMS Track-1 approach is implemented utilizing a detailed network developed in Sub-TAZ approach. The conventional four-step, Sub-TAZ, and TRANSIMS models were estimated in a small case study for Fort Meade, Maryland, with zonal trip tables. The models were calibrated and validated for the base year (2005), and the forecasted results for the year (2010) were compared to actual ground counts of traffic volume and speed. The study evaluated the forecasting ability of TRANSIMS versus the conventional and enhanced four-step models and provided critical observations concerning strategies for the further implementation of TRANSIMS.BACKGROUND Traffic pattern prediction is necessary for infrastructure improvement, and travel demand modeling provides tools to forecast travel patterns under various conditions. This modeling involves a series of mathematical equations that represent how people make travel choices. Traditional travel demand models use the four-step method, which was introduced in the 1950s and has been used widely in transportation planning. Although the four-step method has been practical in producing aggregate forecasts, it has some shortcomings. For example, in short-range planning networks, existing and newly constructed roads become congested much faster than forecasted (TRB 2007) and the performance of current four-step models is not always satisfactory. Additionally, these models are not behavioral in nature and as a result they are unable to represent the time chosen for travel, travelers’ responses to demand policies (e.g., toll roads, road pricing, and transit vouchers), non-motorized

Downloads

Published

2014-03-01

Issue

Section

Articles