Measuring health of highway network configuration against dynamic Origin-Destination demand network using weighted complex network analysis.

Tak, Sehyun; Kim, Sunghoon; Byon, Young-Ji; Lee, Donghoun; Yeo, Hwasoo · 2018 · DOAJ

DOI: 10.1371/journal.pone.0206538

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the need for a systematic method to evaluate the "health" or fitness of highway network configurations against actual, dynamic travel demands. Highway infrastructure represents a permanent, expensive investment, yet previous research often analyzed network topology in isolation from traffic flow or relied on static annual traffic volumes. The authors argue that because congestion arises from temporal inconsistencies between physical network design and short-term peak demands, a methodology is required to compare structural configurations with dynamic Origin-Destination (OD) demand networks. The goal is to provide state-level guidance for monitoring existing networks and designing new highways that better reflect actual usage patterns. The researchers employed weighted complex network analysis using data from the Korean highway system. They conceptualized four network types: two structural networks (Conceptual Network, representing connectivity; and Physical Network, weighted by lane capacity) and two hypothetical OD networks (Movement Network, representing trip existence; and Volume-Weighted Network, weighted by traffic volume). Data included 2015 structural information (614 nodes, lane counts) and 2013 hourly toll collection records for OD pairs. The analysis focused on deriving degree distributions for these networks in log-log planes to identify statistical properties, specifically testing for power law distributions. The findings reveal that the weighted degree distribution of the Physical Network follows a power law with a coefficient of -1.95, whereas the unweighted Conceptual Network shows a narrower, less representative distribution. For the OD networks, the non-weighted Movement Network did not follow a power law; instead, its in-degree and out-degree distributions were modeled as mixtures of normal distributions correlated with city sizes. However, the Volume-Weighted Network, which accounts for actual traffic volumes, followed a power law distribution. Crucially, the power law coefficient of the Volume-Weighted Network was found to be dynamic, changing throughout the day and week, with weekend distributions showing different coefficients than weekdays due to concentrated travel patterns. The significance of this work lies in the proposed methodology for measuring highway network health by comparing the power law coefficients of the static Physical Network against the dynamic Volume-Weighted Network. This approach allows for the detection of deviations from ideal structural configurations associated with actual demands. By identifying these mismatches, the method offers a comprehensive, system-wide tool for evaluating network performance and guiding the construction of new highways to better align with temporal demand fluctuations, thereby improving infrastructure efficiency and user satisfaction.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• exposure measurement