Bewertung von Verfahren zur Erkennung von Stoerungen im Verkehrsablauf in Theorie, Praxis und Simulation Untersuchung von abschnittsbezogenen neuen Verfahren zur automatischen Stau-und Stoerfallerkennung aufgrund lokaler Messwerte und ihre Wirksamkeit im praktischen Einsatz

The research assignment had the objective of assessing procedures for the automatic recognition of traffic disruptions on autobahns. This was to serve, on the one hand, to support a decision-taking basis for the further use of such procedures (in particular, within the framework of the traffic control) and, on the other hand, to obtain findings for the continued optimisation of the procedures. The assessment was composed of system analysis, an empirical field trial and a simulation-aided investigation. Special emphasis was placed on the question of the extent to which new, in particular section-related procedures for the recognition of disruptions are superior to the conventional procedures with regard to suitable quality criteria. The recognition quality of a procedure within the framework of the traffic control is simultaneously oriented towards sensitivity and specificity, i.e. the detection rate compared with the false-alarm rate. The direct determination of the false-alarm rate relies on independent evidence of the existence of a disruption. Within the framework of the investigation approach, methods were developed in order to improve the estimation of the false-alarm rate - although no corresponding evidence exists. One new feature in this investigation was the differentiation of the recognition quality depending on the alarm phase and on an indicator of the degree of the disruption (on the level of a measured drop in speed). These dependences are decisive for the continued optimisation of the procedures with regard to the adjustable threshold values. Within the framework of the system analysis, the concepts of the procedures were analysed and classified theoretically. (orig.)Summary in FrenchAvailable from TIB Hannover: ZA 4681(797)+a / FIZ - Fachinformationszzentrum Karlsruhe / TIB - Technische InformationsbibliothekSIGLEDEGerman