Research")?> Traffic is a high-dimensional dynamic system. The traffic volume varies over day, week and year and also depends on special events, temporary road works, route diversions, etc.. This results in high traffic load fluctuations on our streets. It is not possible to map these dynamics by temporal low dissolved or static models, which are nowadays used for traffic planning .
Even mapping the actual state of traffic is not practicable in general due to the only selective data collection using inductive loops or infrared detectors. From this follows the lack of effectiveness of conventional control strategies, which merely react upon the motoring environment at a time when the traffic jam already exists. A more predictive strategic planning of local and long-distance traffic is necessary to reduce the load of people and environment, especially during times of traffic jam increased pollution.
The last years witnessed the growth of simulational mapping of motoring environment into an efficient tool for the estimation traffic states and predictive evaluation of control actions. The most important working fields of strategic traffic planning are:
annual report 1995/96
annual report 1997/98
annual report 1999/2000
(sorry, only available in german language) you will find an article about our traffic simulation studies.
While working on our project, problems arise that require thoroughly mathematic or scientific research. Therefore look at our topic pool if you are interested in doing your diploma or doctoral thesis at the Center for Parallel Computing.
More traffic")?> For detailled information about our traffic simulation please click:
Literature and Links.