The paper aims to incorporate drivers' perception band of travel time information in en-route diversion models and analyse the net increase in drivers' benefits from the travel time information provision. The paper begins with a review of previous research on the modelling of drivers behaviour in response to traffic information. This is followed by a description of an interactive, computer-based stated preference survey which was designed to offer dynamic en-route travel time information based on the distribution of historic travel time information. The paper then presents the distribution of drivers' perception band of travel time information and discusses the estimation results for the models which incorporate the perception estimation results for the models which incorporate the perception band. Finally the paper demonstrates the possibility to use the calibrated model to measure the net increase in drivers' utility from traffic information provision. It is found that a majority of drivers perceive the travel time information with a perception band of 5 or 10 minutes. The perception band tends to be smaller as the accuracy of information increases. The estimation results for the models indicate that the proposed model incorporating drivers' perception band of travel time information is superior to the model based on the provided travel time information, and it may be possible to be used for assessing driver benefits (the net increase in drivers' utility) from information provision.
Abstract