Empirical Observations of Congestion Propagation and Dynamic Partitioning with Probe Data for Large-Scale Systems
Yuxuan Ji, Jun Luo, Nikolas Geroliminis
Abstract
Yuxuan Ji, Jun Luo, Nikolas Geroliminis
Abstract
Research on congestion propagation in large urban networks has been based mainly on microsimulations of link-level traffic dynamics. However, both the unpredictability of travel behavior and the complexity of accurate physical modeling present challenges, and simulation results may be time-consuming and unrealistic. This paper explores empirical data from large-scale urban networks to identify hidden information in the process of congestion formation. Specifically, the spatiotemporal relation of congested links is studied, congestion propagation is observed from a macroscopic perspective, and critical congestion regimes are identified to aid in the design of peripheral control strategies. To achieve these goals, the maximum connected component of congested links is used to capture congestion propagation in the city. A data set of 20,000 taxis with global positioning system (GPS) data from Shenzhen, China, is used. Empirical macroscopic fundamental diagrams of congested regions observed during propagation are presented, and the critical congestion regimes are quantified. The findings show that the proposed methodology can effectively distinguish congestion pockets from the rest of the network and efficiently track congestion evolution in linear time 0(n).
OpenAlex reports 105 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Research on congestion propagation in large urban networks has been based mainly on microsimulations of link-level traffic dynamics. However, both the unpredictability of travel behavior and the complexity of accurate physical modeling present challenges, and simulation results may be time-consuming and unrealistic. This paper explores empirical data from large-scale urban networks to identify hidden information in the process of congestion formation. Specifically, the spatiotemporal relation of congested links is studied, congestion propagation is observed from a macroscopic perspective, and critical congestion regimes are identified to aid in the design of peripheral control strategies. To achieve these goals, the maximum connected component of congested links is used to capture congestion propagation in the city. A data set of 20,000 taxis with global positioning system (GPS) data from Shenzhen, China, is used. Empirical macroscopic fundamental diagrams of congested regions observed during propagation are presented, and the critical congestion regimes are quantified. The findings show that the proposed methodology can effectively distinguish congestion pockets from the rest of the network and efficiently track congestion evolution in linear time 0(n).
Key concepts: Computer science, Network congestion, Taxis, Traffic congestion, Empirical research, Global Positioning System, Scale (ratio), Process (computing)