An Analytical Travel Time Estimation Method for Planning Models

A travel time estimation method is developed in this study, which takes an analytical form with parameters calibrated based on microscopic simulation data. The developed model has three advantages compared to traditional link-based or node-based models. First, the model considers the influence of signal timing plans for a variety of traffic combinations, although signal timing setting is not explicitly required as input. Second, this model estimates the impacts of queues at different locations upstream of the signal, allowing the splitting of link delays between the controlled link and upstream links based on the link length. Third, the model shows promise of improved travel time prediction accuracy compared to existing models. The testing of the model indicates that mean absolute percentage errors (MAPE) of the model are comparable or better than MAPE of the uniform delay of the HCM 2000 method. The model has the potential for use as part of travel demand forecasting and dynamic traffic assignment models. It could be applied to a large network without the burden of a large amount of signal timing coding, while improving travel time estimation accuracy by implicitly considering signal timing plans.