Using fuzzy clustering of user perception to define levels of service at signalized intersections

Fang Clara Fang, Lily Elefteriadou, Kelly Klaver Pecheux, Martin T. Pietrucha

Research output: Contribution to journalArticle

18 Scopus citations

Abstract

In the current Highway Capacity Manual (HCM) 2000, level of service (LOS) is defined as a quality measure describing operational conditions within a traffic stream. It is also stated in the HCM that "each level of service represents a range of operating conditions and the driver's perception of those conditions." However, the drivers' perception of traffic operational conditions has not been explicitly considered in defining the LOS categories. This study develops and implements a methodology based on fuzzy clustering to define LOS boundaries based on user perception. The methodology is implemented for determining appropriate LOS categories at signalized intersections. These categories are also compared to the existing HCM 2000 LOS categories. An existing database was used containing user perception-related data on estimated delays and ratings of LOS for signalized intersections, and a fuzzy c-means clustering technique was employed to partition the data sets of the estimated delays and of the quality-of-service ratings into LOS categories. It was found that, based on the fuzzy clustering results, each estimated delay is classified into one major LOS category with one minor LOS category. Conclusions drawn from this study are that (1) users can perceive six levels of service in a fuzzy domain in terms of drivers' time-estimating capabilities; (2) the current HCM LOS categories are appropriate for signalized intersections in terms of category range and number; and (3) revision of the HCM LOS definitions with consideration of fuzzy clustering would accommodate subjective variations and uncertainties in user perception and alleviate the existing concerns related to the rigid LOS boundaries.

Original languageEnglish (US)
Pages (from-to)657-663
Number of pages7
JournalJournal of Transportation Engineering
Volume129
Issue number6
DOIs
StatePublished - Nov 1 2003

All Science Journal Classification (ASJC) codes

  • Civil and Structural Engineering
  • Transportation

Fingerprint Dive into the research topics of 'Using fuzzy clustering of user perception to define levels of service at signalized intersections'. Together they form a unique fingerprint.

  • Cite this