A Cellular Automata Framework for Studying Expandable Traffic Flow Models

摘要

The introduction of methods from statistical physics for traffic flow modelling, such as cellular automata models, enabled the faster than real time simulation of large traffic networks. A key issue faced by such efforts is the expandability of the network model. This paper discusses the development of a cellular automata framework for micro-simulation of vehicle traffic flow in road networks and highways, which was implemented on MATLAB. Attention has been paid to the modularity of the proposed model, in order to provide not only expandability, but also parameterization and independency from simulation rules. Our goal is to accomodate the study of traffic critical properties by including several parameters in the model and facilitate the construction of hierarhical modular traffic models and the experimentation with alternative simulation rulesets. The first step of this integrated methodology is to validate the proposed framework, by presenting numerical results and comparisons.