Method for accuracy assessment of aggregated freeway traffic data

Accurate traffic data are crucial for advanced traveller information and traffic management applications. However, it is widely acknowledged that limitations in the monitoring and communication technologies lead to significant inaccuracies in the estimates of traffic parameters such as traffic volume and travel speed. Currently deployed traffic data quality control methods do little more than identifying implausible outliers. This study presents an efficient method based on the concept of spatial consistency for assessing the accuracies of aggregated segment-wise traffic volumes in a freeway network. A novelty of the proposed approach is that unlike most other comparable solutions, it does not depend on the availability of good quality historical traffic data or well-calibrated reference stations. The method has been validated using simulated data generated by a microscopic traffic simulator. It has been found that the accuracy estimates produced by the proposed method strongly correlate with the actual accuracies of the input traffic data that have been corrupted with two types of synthetic errors.