Integrating the Directional Effect of Traffic into Geostatistical Approaches for Travel Time Estimation

Daoqin Tong, Wei Hua Lin, Alfred Stein

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

With the direct linkage to a travel map system, geostatistical techniques have been recently adopted for urban travel time estimation. Some important traffic characteristics of urban transportation networks, however, have not been adequately addressed in these studies. As an improvement over the existing studies, this study incorporates the directional effect of traffic into several commonly used geostatistical models for travel time estimation. We show that model performance can be significantly enhanced when flow specific properties are explicitly considered in constructing the associated interpolation models. The developed methodology is applied to a set of traffic data collected in the city of Tucson, Arizona during the rush hours. Results demonstrate an average of 20 % reduction in RMSE compared with those by the traditional approaches.

Original languageEnglish (US)
Pages (from-to)101-112
Number of pages12
JournalInternational Journal of Intelligent Transportation Systems Research
Volume11
Issue number3
DOIs
StatePublished - Sep 1 2013
Externally publishedYes

Fingerprint

Travel Time
Travel time
Traffic
Urban transportation
Transportation Networks
Performance Model
Linkage
Interpolation
Interpolate
Methodology
Model
Demonstrate

Keywords

  • Directional effect
  • Geostatistical models
  • Travel time estimation
  • Variogram

ASJC Scopus subject areas

  • Neuroscience(all)

Cite this

Integrating the Directional Effect of Traffic into Geostatistical Approaches for Travel Time Estimation. / Tong, Daoqin; Lin, Wei Hua; Stein, Alfred.

In: International Journal of Intelligent Transportation Systems Research, Vol. 11, No. 3, 01.09.2013, p. 101-112.

Research output: Contribution to journalArticle

@article{84f3a80b5c1c470983eb6e69b511647a,
title = "Integrating the Directional Effect of Traffic into Geostatistical Approaches for Travel Time Estimation",
abstract = "With the direct linkage to a travel map system, geostatistical techniques have been recently adopted for urban travel time estimation. Some important traffic characteristics of urban transportation networks, however, have not been adequately addressed in these studies. As an improvement over the existing studies, this study incorporates the directional effect of traffic into several commonly used geostatistical models for travel time estimation. We show that model performance can be significantly enhanced when flow specific properties are explicitly considered in constructing the associated interpolation models. The developed methodology is applied to a set of traffic data collected in the city of Tucson, Arizona during the rush hours. Results demonstrate an average of 20 {\%} reduction in RMSE compared with those by the traditional approaches.",
keywords = "Directional effect, Geostatistical models, Travel time estimation, Variogram",
author = "Daoqin Tong and Lin, {Wei Hua} and Alfred Stein",
year = "2013",
month = "9",
day = "1",
doi = "10.1007/s13177-013-0061-0",
language = "English (US)",
volume = "11",
pages = "101--112",
journal = "International Journal of Intelligent Transportation Systems Research",
issn = "1868-8659",
publisher = "Springer Verlag",
number = "3",

}

TY - JOUR

T1 - Integrating the Directional Effect of Traffic into Geostatistical Approaches for Travel Time Estimation

AU - Tong, Daoqin

AU - Lin, Wei Hua

AU - Stein, Alfred

PY - 2013/9/1

Y1 - 2013/9/1

N2 - With the direct linkage to a travel map system, geostatistical techniques have been recently adopted for urban travel time estimation. Some important traffic characteristics of urban transportation networks, however, have not been adequately addressed in these studies. As an improvement over the existing studies, this study incorporates the directional effect of traffic into several commonly used geostatistical models for travel time estimation. We show that model performance can be significantly enhanced when flow specific properties are explicitly considered in constructing the associated interpolation models. The developed methodology is applied to a set of traffic data collected in the city of Tucson, Arizona during the rush hours. Results demonstrate an average of 20 % reduction in RMSE compared with those by the traditional approaches.

AB - With the direct linkage to a travel map system, geostatistical techniques have been recently adopted for urban travel time estimation. Some important traffic characteristics of urban transportation networks, however, have not been adequately addressed in these studies. As an improvement over the existing studies, this study incorporates the directional effect of traffic into several commonly used geostatistical models for travel time estimation. We show that model performance can be significantly enhanced when flow specific properties are explicitly considered in constructing the associated interpolation models. The developed methodology is applied to a set of traffic data collected in the city of Tucson, Arizona during the rush hours. Results demonstrate an average of 20 % reduction in RMSE compared with those by the traditional approaches.

KW - Directional effect

KW - Geostatistical models

KW - Travel time estimation

KW - Variogram

UR - http://www.scopus.com/inward/record.url?scp=84884145331&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84884145331&partnerID=8YFLogxK

U2 - 10.1007/s13177-013-0061-0

DO - 10.1007/s13177-013-0061-0

M3 - Article

AN - SCOPUS:84884145331

VL - 11

SP - 101

EP - 112

JO - International Journal of Intelligent Transportation Systems Research

JF - International Journal of Intelligent Transportation Systems Research

SN - 1868-8659

IS - 3

ER -