Analysis of traffic statics and dynamics in a signalized double-ring network: A Poincar\'{e} map approach

TL;DR

This paper introduces a novel analytical approach using Poincaré maps to study the static and dynamic behavior of traffic in a signalized double-ring network, revealing stability properties and phenomena like gridlock.

Contribution

It provides the first closed-form analytical framework for analyzing traffic statics and dynamics in signalized double-ring networks, enhancing understanding of network behavior.

Findings

Stationary states can be stable, Lyapunov stable, or unstable.

Multivaluedness and gridlock phenomena are observed on MFDs.

Stationary states are similar for short and long cycle lengths.

Abstract

Understanding traffic statics and dynamics in urban networks is critical to develop effective control and management strategies. In this paper, we provide a novel approach to study the traffic statics and dynamics in a signalized double-ring network, which can provide insights into the operation of more general signalized traffic networks. Under the framework of the link queue model (LQM) and the assumption of a triangular traffic flow fundamental diagram, the signalized double-ring network is studied as a switched affine system. Due to periodic signal regulations, periodic density evolution orbits are formed and defined as stationary states. A Poincar\'{e} map approach is introduced to analyze the properties of such stationary states. With short cycle lengths, closed-form Poincar\'{e} maps are derived. Stationary states and their stability properties are obtained by finding and analyzing the fixed points on the Poincar\'{e} maps. It is found that a stationary state can be asymptotically stable, Lyapunov stable, or unstable. The impacts of retaining ratios and initial densities on the macroscopic fundamental diagrams (MFDs) and the gridlock times are analyzed. Multivaluedness and gridlock phenomena as well as the unstable branch with non-zero average network flow-rates are observed on the MFDs. With long cycle lengths, fixed points on the Poincar\'{e} maps are solved numerically, and the obtained stationary states and the MFDs are very similar to those with short cycle lengths. Compared with earlier studies, this paper provides an analytical framework that can be used to provide complete and closed-form solutions to the statics and dynamics of double-ring networks. This can lead to a better understanding of how the combination of signalized intersections and turning maneuvers is expected to impact network properties, like the MFD.