Development and test of dynamic congestion pricing model

Abstract

Dynamic congestion pricing is an approach to control the traffic flow on the network by setting variable tolls that are adjusted with time based on the traffic condition. Different models have been developed and tested in the past. However, most of these models are based on deterministic network equilibrium rather than stochastic choices of travelers, and case studies on complex networks are rare. These disadvantages limit the use of existing models. In this research, a new dynamic congestion pricing model is developed based on a discrete choice framework to capture users' personal choices. Several solution algorithms are examined and tested in a synthetic network for solving this model. Among these algorithms, SPSA is found to be the best, and is applied successfully to a real case study in Lower Westchester County in New York State. The usefulness and effectiveness of dynamic congestion pricing is also examined and discussed. The results show that dynamic congestion pricing has the potential to improve network performance.