Microscopic Models for Trac Dynamics

• Main Author: Abrahamsson, Felix
• Format: Others
• Language: English
• Published: KTH, Skolan för teknikvetenskap (SCI) 2016
• Online Access: http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-193629

As the volume of vehicular trac has been increasing ever since its advent in the early 20th century, trac dynamics has become a popu- lar topic of study amongst physicists and mathematicians. This paper aims to explain di erent ways of modeling and simulating trac and trac dynamics. In general, modeling of trac can be divided into two types of models, microscopic and macroscopic. In this paper, pri- mary focus is on microscopic models. Examples are shown on how to implement simulations of so called car-following models as well as models based on cellular automata. Certain problems and scenarios regarding trac are studied as well. These include how to eciently distribute a roadblock, how to set the green/red time periods of a traf- c light in order to achieve a high trac ow or a low vehicle density, and how trac ow can be maximized with respect to density or other parameters. Results show that simulations using cellular automata models generally compute faster than simulations using car-following models. However, the results obtained from cellular automata models tend to be more dicult to interpret and apply to real-world sce- narios. A high level of stochasticity in the cellular automata models was also found to be necessary for the models to give applicable re- sults. Car-following models, on the other hand, were found to have the advantage of being more deterministic.