Calibration and Application of a Simulation- Based Dynamic Traffic Assignment Model

The paper presents the calibration and application of a simulation-based dynamic traffic assignment (DTA) model on a portion of the City of Calgary road network in Alberta, Canada. The DTA model iteratively re-assigns flow to paths using the Method of Successive Averages (MSA), based on travel times obtained with a traffic simulation model. The original sub-network extracted from a regional planning model, was enriched by greatly increasing the number of zones. The DTA origin-destination (O-D) matrix was estimated from an extensive database of turning movement counts via a trip generation/distribution model and a matrix adjustment algorithm. The network topology was enhanced by the addition of an interchange and a more precise representation of arterial intersections, including traffic signal control plans. A set of one-hour turning counts was used to calibrate the DTA model by adjusting local parameters such as gap-acceptance values, as well as global parameters such as average vehicle length. The final model results were compared to an independent set of 15-minute turning-movement counts. The resulting R2 values, which ranged from 0.91 to 0.96, lead to a high degree of confidence in the model results