A Markov Process Inspired Cellular Automata Model of Road Traffic

TL;DR

This paper introduces a Markov process-based cellular automata model for road traffic that accurately captures vehicle gap variations and driving behaviors, validated through freeway and intersection scenarios.

Contribution

It presents a novel Markov-Gap cellular automata model that better represents driver behavior and vehicle gaps compared to existing models.

Findings

Model accurately matches empirical gap distributions

Flexible to different traffic scenarios like freeways and intersections

Simulation results align well with observed data

Abstract

To provide a more accurate description of the driving behaviors in vehicle queues, a namely Markov-Gap cellular automata model is proposed in this paper. It views the variation of the gap between two consequent vehicles as a Markov process whose stationary distribution corresponds to the observed distribution of practical gaps. The multiformity of this Markov process provides the model enough flexibility to describe various driving behaviors. Two examples are given to show how to specialize it for different scenarios: usually mentioned flows on freeways and start-up flows at signalized intersections. The agreement between the empirical observations and the simulation results suggests the soundness of this new approach.