Study On The Reliability Of Urban Transportation Networks

Though reliability analysis has been extensively employed in many systems (e.g., power systems, water supply systems, communication networks, etc.), very little attention has been given to reliability analysis of road networks. In fact, good development of the economies heavily depends upon reliable transportation systems. Especially with the advent of Intelligent Transportation System (ITS), demands for better and more reliable services are increasing. However, there are numerous stochastic events (e.g., earthquakes, traffic congestion, accidents, bad weather, road maintenance, etc.) that disrupt the transportation system. Many of these events are unavoidable. Reliability analysis is a prerequisite to minimize the disruption that such events cause. Urban transportation is the critical part of the whole transportation system, so the importance of studying its reliability cannot be overemphasized.On the basis of existing road network reliability and relevant study, this dissertation makes an in-depth study of the urban transportation network reliability from the aspects of route choice behavior under uncertain and road network reliability in the context of information. The main contents contained in this dissertation are as follows:The route choice behavior under uncertain environment is analyzed. Two different route choice models are set up to describe two possible route choice behaviors in the face of uncertain travel times. In the first model users' generalized travel cost are assumed to be weighted sum of the standard deviation and the mean travel time, and users have perfect knowledge of the road network condition. The different weights reflect the travelers' different attitude towards the risk. In the second model, users' route choice decisions are assumed to be tradeoff between travel time and travel time reliability and users only have partial knowledge of the road network condition. The similarities of the two kinds of models are that they both revise the existing models through introducing uncertainty into it. So existing mature algorithms can be used to solve them.The mixed route choice behaviors and corresponding route choice model under the influence of traffic information are discussed. Because mixed...