Abstract
Most studies on travel time reliability (TTR) for roads are single-route section basis due to the availability of sufficient number of travel time observations. Recent years, there have been increasing possibilities of observing network-wide traffic states for long time period with the rapid accumulations of rich probe datasets and it is expected to evaluate TTR for area-wide level rather than route/section levels. This study proposed TTR evaluation method for general space-time domains by using total distance traveled and total travel time that are aggregated based on the generalized definitions of traffic state. We demonstrated the validity and the applicability of the methodology using national level probe dataset by focusing on cross-spatial and cross-seasonal differences or effects of traffic regulations in area-wide TTRs.
