Day-to-day Dynamics In Stochastic Transportation Network With Bounded Rationality

The individuals’ day-to-day route choice behavior is a long time learning and gaming process under the combined effect of historical experience and guidance information, which leads to the day-to-day dynamic evolution of network traffic flow. Individuals continually learn about the travel times from experience or guidance information in a network as they make repeated route choices day-to-day. And the day-to-day dynamics is the aggregate of individual drivers’ route choice.This dissertation analizes the behavioral process of trip makings, establishes a day-to-day dynamic model based on Cumulative Prospect Theory with the consideration of bounded rationality and reference dependency. The effects of the accuracy of guidance information and learning mechanisms on the day-to-day dynamic traffic flow patterns are further examined. In addition, an initial exploration of dynamic traffic assignment considering heterogeneous decision rules is provieded.Firstly, the individual’s behavioral process of urban trips is analysed based on the framework of Cumulative Prospect Theory. And we introduce the behavioral science theory to the study of dynamic traffic assignment and establish a concept structrue of day-to-day adjustment model under guidance information in stochastic network.Secondly, we propose a day-to-day dynamic model to describe drivers’ behavioural changes in urban network with link capacity degradation. The proposed model describes how travellers adjust their travel time budgets as well as route choices from day-to-day according to the updated experience and guidance information from the Reliable Path Searching System. Aditionaly, the role of the accuracy of guidance information and learning mechanisms in driving system-wide properties of traffic network flows over time is also examined.Finally, an initial exploration of heterogeneous decision rules is provided. Traveles are divided into two classes, i.e., in and without compliance with the travel gruidance information system. And two different schemes are presented to describe how the two traveler classes adjust their route choices and the day-to-day dynamics, respectively. Our findings demonstrate that it is essential to give greater emphasis on travelers’ heterogeneity and explore more reasonable ATIS advice according to the specific travel situation and the character of the individual traveler in transport policy and performance management.