An Iterative Procedure for the Estimation of Dynamic Toll Demand, Toll, and Level-of-Service on Toll Facilities

Increasing congestion in the urban highway system, combined with limited funds for new construction has lead to new managed lane strategies designed to provide improved level of service at a lower cost. One of the most popular strategies for managed lanes is dynamic tolling of single-occupant vehicles combined with preferential treatment of high-occupancy vehicles, known as High-Occupancy Toll (HOT) lanes. In order to evaluate these types of strategies, dynamic toll modeling is used to model the relationship of toll demand and toll pricing within different time periods, or more specifically, at different levels of congestion. In this paper, an assignment-based toll choice routine is designed to estimate toll demand within the context of a traditional 4-step travel demand model. This approach uses a dynamic toll demand estimation embedded within an equilibrium highway assignment, assuming the toll choice is a path choice rather than a mode choice. In application, an iterative procedure was employed, involving a modification of the assignment routines for each hour. In each iteration, the toll is adjusted according to the level of service, while the likelihood of a traveler choosing a toll path and ultimately the toll demand is estimated based on the marginal cost saving between toll path and non-toll path. The results of this routine are toll and non-toll trip tables, with associated toll values. A case study is shown to explain how this routine is applied to a toll study where toll and toll demand vary by level of service on toll segments.