Modelling Disrupted Transport Network Behaviour

Aerial photograph of road

Abstract

Current transport modelling tools, utilised for maintenance and development of road infrastructure, assign traffic throughout a network based on traditional equilibrium concepts. These models assume rationality, perfect network knowledge and path cost minimisation of users. Disruptions on a road network such as traffic incidents transform normal driving behaviour and route choice. Hence, these equilibrium concepts cannot appropriately cater for disruptions which restrict the value of the available tools.
This thesis presents an investigation into driving behaviour under disrupted conditions and the subsequent development of traffic assignment models that account for disrupted conditions. It is postulated that disruptions lead to the release of information which results in adaptive behaviour. Information can be through visual cues of the disruption itself, or provided through navigation technology and Advanced Traveller Information Systems (ATIS). Thus, the thesis focuses on two components; impacts of information on driving behaviour and the incorporation of information into network modelling.
The concepts of experimental economics were used to develop two controlled laboratory experiments; the first studied the presence of information on risk attitudes of users. The findings revealed that users became risk neutral in the presence of information resulting in the overvaluation of information sources. The second experiment empirically proved the existence of an online information paradox where the provision of information results in deterioration of network performance. Neglecting these findings may affect the economic appraisal of informational infrastructure and can potentially result in erroneous planning outcomes.
Principles of “User Equilibrium with Recourse” (UER) were used to develop static (Disrupted Equilibrium Assignment with Recourse (DEAR)) and dynamic (Dynamic User Optimal with Recourse (DUOR)) traffic assignment methodologies which account for the adaptive behaviour of users in light of a disruption. Application of the models suggests an increase in system wide costs when accounting for disruptions and highlights the benefits of information provision.
The core contribution of the research is the development of novel modelling techniques that incorporate adaptive behaviour associated with disruptions. Further development of these models can lead to improved assessment of reliability and vulnerability of road networks.

Reference

WIJAYARATNA, K. (2016). Modelling Disrupted Transport Network Behaviour. Civil & Environmental Engineering, Faculty of Engineering, UNSW. (2016). Web.

Registrations Open - August, October and December 2021 Online Training. View Upcoming Events..