A Traffic Cellular Automaton with Estimation of Time to Collision
Yohei Taniguchi and Hideyuki Suzuki
We present a new traffic CA model in which each vehicle moves considering an estimated time for a following vehicle to catch up with the one ahead (Time-to-Collision: TTC), and investigate characteristics of the model with the simulation. We also analyze analytically the possibility of a collision between two cars in this model. The model is simulated under open boundary conditions and each car is parallel-updated. We show fundamental diagrams of the traffic flow in the simulation. In the diagrams, we find two distinct phases: a free flow and wide moving jam. Between the two phases, the region where the dots spread sparsely is seen clearly. In addition to this, by using different values of parameter, we can see the several patterns of the trajectory of vehicles. Based on these findings, we believe it is possible for this model to reproduce synchronized flow.
Keywords: Traffic flow, TTC, time-to-collision, CA, cellular automaton, synchronized flow