Deriving the Value of Time for Toll Revenue Forecasting using Multiple Criteria Decision-Making

Abstract

Mathematical traffic models replicate travel decisions and driver behavior. These tools have become more sophisticated, yet they remain simplified representations of more complex systems, including associated toll forecasts. A literature review shows that a substantial amount of toll revenue forecasts has fallen short of accurately representing actual road usage—by as much as 50%. In recent years, simulation-based modeling tools have emerged as an innovative approach to regional traffic forecasting. They capture time-varying traffic conditions and represent more realistic traffic flow and congestion, but this type of travel forecasting new and complex. One key variable in these traffic modeling tools is the value of time (VoT). The VoT represents how drivers perceive their willingness to pay a toll to reduce their travel time. How to properly quantify the VoT is still not completely understood, but it is the governing factor in how toll revenue forecasts are calculated. We propose three different approaches for quantifying the VoT. The first approach uses socio-economic data in terms of median income per zone. The VoT is quantified based on income levels of each origin-zone that travelers depart from. The second approach uses a traveler’s destination zone—derived based on trip purpose. The second approach uses a multiple-criteria decision-making (MCDM) methodology known as the Analytical Hierarchy Process (AHP) to develop weighting for each trip purpose. A third approach also uses AHP (combined with departure time) to derive weights for each traveler’s VoT. The three approaches were tested using a simulation-based dynamic traffic assignment (DTA) model.