Abstract: 

Congestion in capacity-constrained urban environments is essentially the aggregate outcome of choices that individuals make in self-interest while assuming that others do the same. Existing models in transportation do typically no take the interaction and dependencies between different travellers into account. In this study, a new modelling approach was developed that incorporates an endogenous interaction term in a discrete choice model to describe the strategic choice behaviour of individuals. The approach accounts for different degrees of rationality, and choice situations, reflecting the extent of previous experiences. The choice models are linked to a model of how a control agent acts strategically, dependent on different objectives in optimizing usage of capacity-constrained facilities. The paper summarizes the approach, discusses the results of numerical simulations to assess its face validity and reports some main findings from two experiments that were conducted to empirically estimate the models.