The optimization of traffic systems by expanding or improving an existing network is often referred to as a Network Design Problem (NDP). The optimization of dynamictraffic management (DTM) measures can be seen as a variant of the NDP. Traditionally, this type of optimization is focused on improving accessibility, given particular boundary conditions for traffic safety and ivability (set by law). Yet, the qualityoftraffic systems is not only a matter of accessibility, but also of externalities such as traffic safety, air quality and climate. This paper describes the results of research on optimization of DTM measu res to improve externalities. The formulation and calculation of the objective functions are described using a dynamic traffic assignment (DTA) model and three solution approaches are presented using genetic algorithms and compared in a case study. Furthermore, an analysis of objectives being opposïte or aligned when a traffic system is being optimized using these measures is presented. Within this research the NDP problem is formulated as a bi-level optimization problem. The objectives for congestion, traffic safety, emissions and noise show different optimal solutions,which means there is not a combination of measures resulting in an optimal situation for a combination of all objectives. However; it was found that the objectives concerning emissions and congestion are aligned and that these objectives are opposite to noise and traffic safety. This shows that it is not true that improving accessibility will also improve externalities and that finding solutions for decreasing of multiple externalities is not easy. A public policy decision is needed for choosing the solution. (Author/publisher)
Abstract