Empirical validation of a queueing approach to uninterrupted traffic flows (Q862815)

The paper reports an empirical validation of the validity of queueing models for modeling uninterrupted highway traffic flows. A dataset collected by the ministry of transportation of the Flemish Government (Belgium) was used in the analysis. Since traffic flows change during the day, different queueing models were used for different periods of the day. The utilized approach was to find a queueing model with a small difference between the observed speed and the speed predicted by a queueing model. The selected measure of the difference was the Theil coefficient. The conclusion of the study is that during non-congested hours the best model is \(M/G/1\). During the congested hours, the state dependent \(GI/G/z\) models are more realistic. For \(GI/G/z\) models the authors use approximate formulae.