This thesis focuses on building up a systematic framework that is capable of analyzing the ability of road networks against unpredictable and exceptional incidents, such as a severe accident. This ability can be simply treated as the road network robustness. For the performance analysis on a network scale, it is crucial to represent the route choice behavior of travelers and dynamic traffic assignment (DTA). Starting from the analysis of the basic requirements for the complete study of road network robustness, an innovative two-step process is proposed. It's basic idea is that in step one an user equilibrium status of the network is built up to represent the daily traffic pattern; and in step two, based on the equilibrium status, incident scenarios are represented as a non-equilibrium situation or a new equilibrium situation taking into account the effects of the incident. So the framework is also designed with two stages corresponding to these two steps, and our main contribution is to adopt two approaches of DTA models (i.e., user equilibrium (UE) approach and en-route approach) into the framework. Through case studies of two road networks with different size and complexity, the applicability and effectiveness of this framework have been proved for the systematical analysis of a road network robustness against different types of incidents. (Author/publisher)
Samenvatting