A new method of abnormal data detection on traffic flow of extra long highway tunnel

Fuhua Yu, Hongke Xu, Qi Wang, Guoqiang Li

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Researched on the characteristic of the traffic flow, a new method of abnormal data detection on traffic flow based on the curve-fitting is presented. First, the historical data on traffic flow are divided into the traffic flow with high speed and the traffic flow with low speed by the clustering analysis. Then, the algorithm on the determination safety zone scope is obtained by the curve-fitting on the historical data and the undetermined abnormal data can be detected from the real-time traffic flow data for out of the calculated safety zone scope. Finally, the undetermined abnormal data are confirmed as the abnormal data which are even beyond the accepted error range. The real-time traffic flow data of an extra long highway tunnel in Shaanxi are taken as examples, and the detection result by the presented method shows that the abnormal data can be efficiently and effectively detected from the real-time data on traffic flow of extra long highway tunnel.

Original languageEnglish
Title of host publication2010 International Conference on Logistics Engineering and Intelligent Transportation Systems, LEITS2010 - Proceedings
Pages59-62
Number of pages4
DOIs
StatePublished - 2010
Externally publishedYes
Event2010 International Conference on Logistics Engineering and Intelligent Transportation Systems, LEITS2010 - Wuhan, China
Duration: 26 Nov 201028 Nov 2010

Publication series

Name2010 International Conference on Logistics Engineering and Intelligent Transportation Systems, LEITS2010 - Proceedings

Conference

Conference2010 International Conference on Logistics Engineering and Intelligent Transportation Systems, LEITS2010
Country/TerritoryChina
CityWuhan
Period26/11/1028/11/10

Keywords

  • Abnormal data
  • Curve-fitting
  • Data detection
  • Safety zone
  • Traffic flow

Fingerprint

Dive into the research topics of 'A new method of abnormal data detection on traffic flow of extra long highway tunnel'. Together they form a unique fingerprint.

Cite this