A comparison of data-fitted first order traffic models and their second order generalizations via trajectory and sensor data
For traffic flow modeling researchers, this paper provides an empirical comparison showing that the added complexity of second-order models yields marginal predictive gains.
The study compares first-order LWR traffic models with second-order ARZ generalizations using trajectory and sensor data, finding that ARZ models do not significantly improve predictive accuracy over LWR models.
The Aw-Rascle-Zhang (ARZ) model can be interpreted as a generalization of the first order Lighthill-Whitham-Richards (LWR) model, possessing a family of fundamental diagram curves, rather than a single one. We investigate to which extent this generalization increases the predictive accuracy of the models. To that end, a systematic comparison of two types of data-fitted LWR models and their second order ARZ counterparts is conducted, via a version of the three-detector problem test. The parameter functions of the models are constructed using historic fundamental diagram data. The model comparisons are then carried out using time-dependent data, of two very different types: vehicle trajectory data, and single-loop sensor data. The study of these PDE models is carried out in a macroscopic sense, i.e., continuous field quantities are constructed from the discrete data, and discretization effects are kept negligibly small.