Networks of fixed-cycle intersections

TL;DR

This paper introduces an algorithmic approach to analyze traffic flow in networks of fixed-cycle intersections, capturing wave phenomena and platoon formation, validated against simulations and applicable to complex networks.

Contribution

It extends the fixed-cycle traffic-light queue model to network analysis, enabling decomposition of networks into isolated intersections with accurate modeling of network effects.

Findings

Algorithm matches simulation results accurately.

Captures wave phenomena and platoon formation.

Applicable to complex intersection networks.

Abstract

We present an algorithmic method for analyzing networks of intersections with static signaling, with as primary example a line network that allows traffic flow over several intersections in one main direction. The method decomposes the network into separate intersections and treats each intersection in isolation using an extension of the fixed-cycle traffic-light (FCTL) queue. The network effects are modeled by matching the output process of one intersection with the input process of the next (downstream) intersection. This network analysis provides insight into wave phenomena due to vehicles experiencing progressive cascades of green lights and sheds light on platoon forming in case of imperfections. Our algorithm is shown to match results from extensive discrete-event simulations and can also be applied to more complex network structures.