Citation

Chang H, Park D. Int. J. Urban Sci. 2021; 25(2): 193-207.

Copyright

(Copyright © 2021, University of Seoul, Publisher Informa - Taylor and Francis Group)

DOI

10.1080/12265934.2020.1816206

PMID

unavailable

Abstract

Vehicle-trajectory big data, collected through vehicle-GPS systems, is one of key information sources for road traffic volumes because the vehicle-trajectory volume represents a certain portion of the total traffic volume. This renders a promising opportunity in surveying annual average daily traffic (AADT) volumes. Devising viable means of surveying AADT volumes for all road segments with limited budgets and resources remains one of the main challenges in urban transportation studies. This paper proposes a new methodology for directly surveying AADT volumes using vehicle-GPS trajectory data. The methodology consists of two sub-methods: a nonlinear spatial clustering method based on trip connectivity between observed road sections and a target road section for selecting effective road sections from a road network, and a direct conversion method to understand the nonlinear relationship between the annual average daily probe (AADP) volumes and the observed AADT volumes for the selected road sections, with the subsequent expansion of the AADP volume of the target road section into the AADT volume. In a case study with real-world vehicle-GPS trajectory data, the performance of the method was found to be highly acceptable for actual applications in terms of its estimation accuracy. Therefore, it appears that the proposed method is and will be feasible in the present and near future.

Language: en

Keywords

conversion of vehicle-GPS probe; direct traffic demand survey; nonlinear spatial clustering; trip connectivity; Vehicle-GPS trajectory big data