Hybrid multilane models for highway traffic

TL;DR

This study compares different lane-changing schemes in multilane highway traffic models using cellular automata, finding that an asymmetric 'Keep Right Unless Overtaking' model improves traffic flow efficiency.

Contribution

It introduces and evaluates a hybrid multilane traffic model with asymmetric lane-changing rules, demonstrating its superior efficiency over symmetric and other hybrid schemes.

Findings

Asymmetric model outperforms symmetric and hybrid models in traffic flow.

'Keep Right Unless Overtaking' rule enhances traffic efficiency.

Software package for modeling is freely available.

Abstract

We study effects of lane changing rules on multilane highway traffic using the Nagel-Schreckenberg cellular automaton model with different schemes for combining driving lanes (lanes used by default) and overtaking lanes. Three schemes are considered: a symmetric model, in which all lanes are driving lanes, an asymmetric model, in which the right lane is a driving lane and the other lanes are overtaking lanes, a hybrid model, in which the leftmost lane is an overtaking lane and all the other lanes are driving lanes. In a driving lane vehicles follow symmetric rules for lane changes to the left and to the right, while in an overtaking lane vehicles follow asymmetric lane changing rules. We test these schemes for three- and four-lane traffic mixed with some low-speed vehicles (having a lower maximum speed) in a closed system with periodic boundary conditions as well as in an open system with one open lane. Our results show that the asymmetric model, which reflects the "Keep Right Unless Overtaking" rule, is more efficient than the other two models. An extensible software package developed for this study is free available.