Analysing day-to-day and within-day changes of ramp-to-ramp traffic volume on urban expressway using Gaussian process

Kenji Ikeda, Shohei Yasuda, Takamasa Iryo, Masaaki Ishihara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The analysis of the day-to-day and within-day changes of the ramp-to-ramp traffic volume of an urban expressway is important for efficient operation and management of vehicular traffic. A data-oriented approach is proposed in this investigation to estimate day-to-day and within-day changes of ramp-to-ramp traffic volume using three types of traffic data: ramp-to-ramp traffic volume, travel time between ramps, and section traffic volume data of arterial roads. In this regard, a negative binomial regression model and a Gaussian process regression model were developed and their accuracy was evaluated. It was determined that the latter generally yielded better accuracy and its characteristics were examined.

Original languageEnglish
Title of host publicationProceedings of the 24th International Conference of Hong Kong Society for Transportation Studies, HKSTS 2019
Subtitle of host publicationTransport and Smart Cities
EditorsAndy H.F. Chow, S.M. Lo, Lishuai Li
PublisherHong Kong Society for Transportation Studies Limited
Pages99-106
Number of pages8
ISBN (Electronic)9789881581488
Publication statusPublished - 2019
Externally publishedYes
Event24th International Conference of Hong Kong Society for Transportation Studies: Transport and Smart Cities, HKSTS 2019 - Hong Kong, Hong Kong
Duration: 2019 Dec 142019 Dec 16

Publication series

NameProceedings of the 24th International Conference of Hong Kong Society for Transportation Studies, HKSTS 2019: Transport and Smart Cities

Conference

Conference24th International Conference of Hong Kong Society for Transportation Studies: Transport and Smart Cities, HKSTS 2019
Country/TerritoryHong Kong
CityHong Kong
Period19/12/1419/12/16

Keywords

  • Gaussian process
  • Generalized linear model
  • Non-linear model

ASJC Scopus subject areas

  • Building and Construction
  • Information Systems
  • Civil and Structural Engineering
  • Transportation
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Analysing day-to-day and within-day changes of ramp-to-ramp traffic volume on urban expressway using Gaussian process'. Together they form a unique fingerprint.

Cite this