• DocumentCode
    3026047
  • Title

    Expressway Emergency Resources Demand Forecasting Based on Neural Network

  • Author

    Liu Jin

  • Author_Institution
    Dept. of Traffic Eng., Hong Kong-Zhuhai-Macao Bridge Authority, Zhuhai, China
  • fYear
    2013
  • fDate
    29-30 June 2013
  • Firstpage
    595
  • Lastpage
    598
  • Abstract
    Expressway traffic accidents seriously threaten the personal property security. And emergency resources are the basis and premise of accident rescue. Thus the emergency resource demand prediction of expressway is of great significance. The influence factors of emergency resource demand is analyzed in this paper, and neural network programming is carried out on the highway emergency resource demand. Finally, combining trained of neural network and case analysis, it achieves emergency resource demand projections for the new case of the emergency center. The results show that the BP neural network can form the inherent law of highway emergency resource demand after training, self-learning and self-adaptation, and the results can meet the prediction error precision. So the results of neural network prediction can provide scientific allocation of expressway emergency resource with reasonable reference.
  • Keywords
    backpropagation; emergency management; neural nets; road accidents; road safety; road traffic; traffic engineering computing; BP neural network; accident rescue; case analysis; emergency center; emergency resource demand prediction; expressway emergency resource demand forecasting; expressway traffic accident; highway emergency resource demand; neural network prediction; neural network programming; personal property security; self-adaptation; self-learning; Accidents; Biological neural networks; Hazards; Resource management; Roads; demand forecasting; emergency resources; expressway; neural network; traffic engineering;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Digital Manufacturing and Automation (ICDMA), 2013 Fourth International Conference on
  • Conference_Location
    Qingdao
  • Type

    conf

  • DOI
    10.1109/ICDMA.2013.140
  • Filename
    6598061