This paper presents a framework for modeling state dependency in lane changing behaviour of drivers and demonstrates it through an on-ramp merging model for congested freeway situations. The proposed model explicitly considers the anticipation of future conditions as a basis of decision-making and incorporates state dependence to capture the effects of past decisionsthe driver has made on his current decision-making process. The structureof the state-dependent model is described and the estimation data and estimation methodology are presented, followed by results. Improvements in the proposed model are demonstrated by statistical comparisons of the model against an instantaneous model that is estimated with the same dataset ignoring state dependency. For the covering abstract see ITRD E144727. Reprinted with permission of Elsevier.
Samenvatting