Anwendung von Unsicherheitsmodellen am Beispiel der Verkehrserfassung unter Nutzung von Mobilfunkdaten [article]

Ralf Borchers, Universität Stuttgart, Universität Stuttgart
The high and steadily growing volume of traffic is frequently the cause for excessive strains on traffic routes and resulting consequences, such as traffic jams and accidents. The use of traffic management systems aims to achieve improvements in the efficiency of traffic flow on the existing traffic routes. Traffic management systems require traffic information, which can be obtained from induction loops or from so-called floating car data. Since cellular phones can be located these so-called
more » ... d these so-called floating phone data is suitable for traffic detection as well. This method of traffic detection, although economically attractive, is accompanied by a high level of imprecision in the localisation of up to several hundred metres as well as by missing information about whether and in which vehicle the cellular phones were transported. In this paper, mobile positioning is based on a signal level matching method, wherein the position of the cellular phone is determined by use of the received signal strength at the mobile phone and matched to a signal strength map of the network provider. The random and systematic uncertainties are modelled alternatively using random variability, fuzzy theory or fuzzy randomness. In the following identification methods are presented, that indentify mobile phone data generated in motorized individual vehicles (e.g. in cars or trucks). At the beginning, it is checked whether the cellular phone is in motion. If it moves, its velocity is taken to identity the means of transport. Background for this decision is that design-related or administrative reasons limit the speed of some means of transport. Some means of transport can be excluded if significantly higher speeds exist for the cellular phone. As mobile data generated in public transport are unsuitable for the detection of individual vehicles, these are filtered out in the next step. From timetables the positions of the vehicles of the public transport (e.g., busses, trams) are predicted and compared with the positions of the cellular phone both [...]
doi:10.18419/opus-3928 fatcat:g46nqu53drczbgvwd5v3ngg5da