Adaptive strategies for route selection en-route in transportation networks

TL;DR

This paper investigates adaptive route selection strategies in transportation networks using a cellular automata model, revealing how vehicle behavior updates influence traffic flow and congestion management.

Contribution

It introduces a model where vehicles adapt their path-greediness based on local traffic, highlighting optimal update timing and magnitude for improved traffic coordination.

Findings

Optimal update steps depend on vehicle density

Gradual, infrequent updates improve traffic flow in dense networks

Path-greediness adjustments significantly impact congestion levels

Abstract

We examine adaptive strategies adopted by vehicles for route selection en-route in transportation networks. By studying a model of two-dimensional cellular automata, we model vehicles characterized by a parameter called path-greediness, which corresponds to the tendency for them to travel to their destinations via the shortest path. The path-greediness of each individual vehicle is updated based on the local traffic conditions, to either keep the vehicle travels via a shorter path in an un-congested region or to explore longer diverted paths in a congested region. We found that the optimal number of steps to trigger an update of path-greediness is dependent on the density of vehicles, and the magnitude of path-greediness increment affects the macroscopic traffic conditions of the system. To better coordinate vehicles in denser networks, the update on the tendency for vehicles to travel via the shorter paths should be gradual and less frequent.