CELLULAR AUTOMATA FOR HIGHWAY TRAFFIC FLOW SIMULATION.

A Cellular Automaton is a extremely simplified program for the simulation of complex transportation systems, where the performance velocity is more important than the detailed model accuracy. In a Cellular Automaton, space and time are divided into discrete cells and steps. A cell exchanges only transported units with the neighboring cells directly within one time step. A Cellular Automaton model is specified by the rules which control these exchanges. The first application of the Cellular Automaton for simulation of traffic flows on streets and highways was introduced by Nagel and Schreckenberg (1992). The basic Cellular Automaton model from Nagel-Schreckenberg has been checked against measurements of realistic traffic flow on urban streets and motorways in Germany. It was found that the measured capacities on German motorways cannot be reproduced very well. On urban streets, however, it was very well possible to represent traffic patterns at intersections. The paper describes a completely new concept for the cellular automaton principle to model highway traffic flow. This model uses a time-oriented car-following model. This model accounts for the real driving behavior more precisely than the model from Nagel and Schreckenberg. The properties of discrete time and space are maintained. Also the updating process can be done for each vehicle independently from each other. Tests showed that this model achieves rather good correspondence with the observed real traffic flow. This paper shows that a Cellular Automaton is generally applicable for simulation of traffic flows. The degree of correspondence with reality depends on the applied car-following model. The new model concept combines realistic modeling with fast computational performance.