This thesis develops four cell-based dynamic traffic assignment models. Two of them are for the route choice problems with fixed demand. One is for the simultaneous route and departure time choice (SRD) problem with fixed demand and the last one is for the SRD problem with elastic demand. The first route choice model is formulated by the Nonlinear Complementarity Problem (NCP) approach while the other DTA models are formulated through the Variational Inequality (VI) approach. These formulations follow the ideal dynamic user optimal (DUO) principle. To improve the accuracy of dynamic traffic modeling, these formulations encapsulate a network version of the Cell Transmission Model (CTM). Moreover, these formulations satisfy the first-in-first-out (FIFO) conditions through the Cell Transmi...[ Read more ]

This thesis develops four cell-based dynamic traffic assignment models. Two of them are for the route choice problems with fixed demand. One is for the simultaneous route and departure time choice (SRD) problem with fixed demand and the last one is for the SRD problem with elastic demand. The first route choice model is formulated by the Nonlinear Complementarity Problem (NCP) approach while the other DTA models are formulated through the Variational Inequality (VI) approach. These formulations follow the ideal dynamic user optimal (DUO) principle. To improve the accuracy of dynamic traffic modeling, these formulations encapsulate a network version of the Cell Transmission Model (CTM). Moreover, these formulations satisfy the first-in-first-out (FIFO) conditions through the Cell Transmission Model. Thirteen scenarios are set up to evaluate the properties of these formulations, in the areas of traffic dynamics, traffic interactions across multiple links and the ideal DUO principle. The results show that the formulations are capable of capturing dynamic phenomena, such as shock waves, queue formation and dissipation, and traffic interactions, such as queue spill back. Moreover, the results demonstrate that these cell-based formulations produce solutions that precisely follow the ideal DUO principle.