Inter-urban short-term traffic congestion prediction

The main aim of Dynamic Traffic Management is efficient and effective use of the existing traffic infrastructure network. To reach this goal, traffic operators take measures, based on the current traffic situation, control schemes and information services at hand, that (try to) influence and distribute traffic in such a manner that it makes optimal use of the present road infrastructure. In this dissertation, the performances of models that were developed to produce accurate congestion prediction are compared. The prediction horizon is short-term period of 5, 10, 15, 20, 25, or 30 minutes. All assessed models are based on data driven methods.