The author describes the use of section-related data detection for incident and queue detection applications. Obtaining section-related traffic data is based on the idea of deriving a feature sequence from the vehicle series detected by a local measurement stations. Detuning curves are used to characterise individual vehicles and artificial intelligence techniques applied to produce individual feature vectors. A comparison of data from neighbouring stations reveals the occurrence of incidents and their severity. The use of the technique on a motorway between Stuttgart and Waiblingen is described.
Abstract