FREEWAY TRAVEL TIME FORECAST USING ARTIFICIAL NEURAL NETWORKS WITH FEATURE COMPOSITION

DOI
  • Lee Ying
    Department of Hospitality Management, MingDao University
  • Wei Chien-Hung
    Department of Transportation and Communication Management Science, National Cheng Kung University
  • Lan Lawrence W.
    Department of Global Marketing and Logistics, MingDao University

Description

This paper develops a novel travel time forecasting model using artificial neural networks with feature composition. The core logic of the model is based on a functional relation between real-time traffic data as the input variables and actual travel time data as the output variable. Feature composition is employed to reduce the data features with fewer input variables while still preserving the relevant traffic characteristics. The forecasted travel time is then obtained by plugging in the real-time traffic data into the functional relation. To validate the model, sufficient amounts of actual travel time and real-time traffic data are collected from Taiwan freeway intercity buses (equipped with global positioning systems), roadway vehicle detectors and the accident database. The proposed travel time forecasting model has shed some light on the practical applications in the intelligent transportation systems context.

Journal

Keywords

Details 詳細情報について

  • CRID
    1390001205675162112
  • NII Article ID
    130005037043
  • DOI
    10.11175/eastpro.2009.0.310.0
  • Text Lang
    en
  • Data Source
    • JaLC
    • CiNii Articles
  • Abstract License Flag
    Disallowed

Report a problem

Back to top