Metastable States in Cellular Automata for Traffic Flow

TL;DR

This paper introduces a generalized cellular automaton model for traffic flow that exhibits metastable states, providing insights into hysteresis effects observed in real traffic but absent in previous models.

Contribution

A simple generalization of the NaSch model with velocity-dependent randomization that demonstrates metastable states in traffic flow modeling.

Findings

The generalized model exhibits metastable states.

It sheds light on conditions for hysteresis in flow-density relations.

The model aligns better with real traffic phenomena.

Abstract

Measurements on real traffic have revealed the existence of metastable states with very high flow. Such states have not been observed in the Nagel-Schreckenberg (NaSch) model which is the basic cellular automaton for the description of traffic. Here we propose a simple generalization of the NaSch model by introducing a velocity-dependent randomization. We investigate a special case which belongs to the so-called slow-to-start rules. It is shown that this model exhibits metastable states, thus sheding some light on the prerequisites for the occurance of hysteresis effects in the flow-density relation.