Center for Transportation Studies, University of Minnesota
In this paper, we explore the usefulness of cellular automata to traffic flow modeling. We extend some of the ex.isting
CA models to capture characteristics of traffic flow that have not been possible to model using either conventioinal
analytical models or existing simulation techniques. In particular, we examine higher moments of traffic flow and
evaluate their effect on overall traffic performance. The behavior of these higher moments is found to be surpri, ,lg,
somewhat counter-intuitive, and to have important implications for design and control of traffic systems. For
example, we show that the density of maximum throughput is near the density of maximum speed variance. Contrary
to current practice, traffic should, therefore, be steered away from this density region. For deterministic systems we
found traffic flow to possess a finite period which is highly sensitive to density in a non-monotonic fashion. We show
that knowledge of this periodic behavior to be very useful in designing and controlling automated systems. These
results are obtained for both single and two lane systems. For two lane systems, we also examine the relationship
between lane changing behavior and flow performance. We show that the density of maximum lane changing
frequency occurs past the density of maximum throughput. Therefore, traffic should also be steered away from this
Benjaafar, Saifallah; Dooley, Kevin; Setyawan, Wibowo.
Cellular Automata for Traffic Flow Modeling.
Center for Transportation Studies, University of Minnesota.
Retrieved from the University of Minnesota Digital Conservancy,
Content distributed via the University of Minnesota's Digital Conservancy may be subject to additional license and use restrictions applied by the depositor.