• DocumentCode
    3702737
  • Title

    Machine learning for the activation of contraflows during hurricane evacuation

  • Author

    John W. Burris;Rahul Shrestha;Bibek Gautam;Bibidh Bista

  • Author_Institution
    Department of Computer Science and Industrial Technology, Southeastern Louisiana University Hammond, LA, USA
  • fYear
    2015
  • Firstpage
    254
  • Lastpage
    258
  • Abstract
    Contraflows are a critical part of an emergency evacuation plan. In most cases, a contraflow lane reversal will double the capacity of key evacuation routes. The Contraflow plan for the evacuation of southeast Louisiana during a hurricane threat uses a typical schedule for the activation of contraflows based on the predicted time of landfall. This work will apply machine learning techniques using real-time traffic data to schedule the activation of contraflows. Optimizing the Contraflow plan should increase the effectiveness of the evacuation plan by increasing the flow of evacuation traffic based on demand and retaining the availability of incoming traffic until contraflow lanes are needed. These techniques could be applied to other locations, including those without an existing evacuation plan.
  • Keywords
    "Hurricanes","Machine learning algorithms","Delays","Algorithm design and analysis","Supervised learning","Prediction algorithms","Classification algorithms"
  • Publisher
    ieee
  • Conference_Titel
    Global Humanitarian Technology Conference (GHTC), 2015 IEEE
  • Type

    conf

  • DOI
    10.1109/GHTC.2015.7343981
  • Filename
    7343981