Road pricing is a controversial topic gaining high attention from society, not only researchers, but planners, politicians, economists, travelers, etc. Pricing is generally seen as a promising approach to help solving today's transport problems such as congestion, funding of road projects, negative impact of traffic to the environment, etc. Although promising, designing and applying a pricing strategy is a complex issue. Looking at this problem from only one perspective seems not enough to solve the problem because many actors are involved. To support the road authority in optimizing the traffic network performance, a practical tool is needed. A combination of traffic modeling, mathematical models, and control is suitable to give support to traffic engineers in network optimization. Special attention in this thesis is given to differences in travel behavior of different types of travelers, in contrast to most road pricing studies considering homogeneous travelers only. We believe that, especially in the road pricing problem, the behavioral differences between travelers play a crucial role. Hence, our aim in this thesis is to consider explicitly differences in travel behavior and take these into account in the various modeling approaches. The main result of this thesis is the development and application of a design tool (including the developed travel choice models) for optimizing the system set-up (levels of tolls to be levied) of different road pricing regimes in dynamic traffic networks. Moreover, in this thesis we extended a number of notions with respect to tolling to the situation of time-varying tolls to be applied in dynamic traffic networks with dynamic travel demand. This refers to possible policy objectives, road pricing regimes, and road pricing measures. Namely, developing the road pricing model (including different pricing regimes and different policy objectives) adds an additional optimization level which makes the formulation and solution of the dynamic optimal toll design problem more complex. The developed methodology is assessed by applications on small test networks showing that the optimal toll settings can be determined if regimes are given and the travel demand is known. (Author/publisher)
Samenvatting