Estimating travel time reliability in urban areas through a dynamic simulation model

It has been recognised that travelers do not only take travel time into account, but also travel time reliability (TTR). In the presence of travel time unreliability, travellers typically allow more time for their trips in order to limit the possibility of arriving late. This extra time allowance could be reduced increasing transport reliability, with a clear user benefit. Therefore, the evaluation of TTR has been receiving considerable attention in recent years, also in consideration of the current availability of real-time data. This study proposes a methodology for estimating TTR of an extended road network, through the calibration of empirical relations and the finding's representation on GIS maps. The basic variables used are travel times of each link belonging to the network, estimated through a traffic model, combining dynamic assignment with rolling horizon technics and real-time traffic measures from radar sensors. Different statistical measures are used to analyze the quality of the database and its usefulness for monitoring and quantifying TTR, such as Standard Deviation, Coefficient of Variation (CV) and Congestion Index (CI). The methodology is applied to the urban area of Catania (Italy) and three levels of analysis are performed: a Single Link Analysis, a Multi-Link Analysis and a Global Network Analysis. Empirical relations between CV and CI are found and will be used in the appraisal of transport network's performances based on TTR. This work concludes discussing how the process for estimating TTR, as a short term measure, can be attained in urban conditions