Optimising time-varying network flows by low-revenue tolling under dynamic user equilibrium

Abstract

This paper presents a dynamic congestion charging framework that includes a dynamic user equilibrium (DUE) network model to optimize a tolling scheme, whose objective(s) may be determined by the transport planner. The framework presented may be utilised to evaluate potential charging regimes; this paper concentrates on the feasibility of extending the concept of minimal-revenue tolls from the static to time-varying environment. We discuss the limitations in the transference of this optimisation objective and propose a pragmatic approach designed to allow the planner to balance the effect of extracting additional toll revenue against achievable network benefit. We investigate the ability of this framework to incorporate various tolling scenarios and to illustrate the production of low-revenue tolls. The numerical tests use a fixed toll across the entire time period for a range of demand scenarios and a fairly stable tolling solution is observed.