FORECASTING OF AIR POLLUTION IN URBAN AREAS BY MEANS OF ARTIFICIAL NEURAL NETWORKS

This paper addresses the problem of optimum data dimensionality reduction in relation to analysis of weather factors and their influence on the recorded air pollution concentrations in urban areas. Eight weather parameters and two air pollution factors, concentrations of sulfur oxide and suspended particulate are considered. Principal component analysis is used for an optimum decorrelation and dimensionality reduction of the analyzed weather factors. A method for prediction of the concentration of suspended particulate and SO2 based on the artificial neural networks was developed with a possibility to forecast the pollution level one day in advance.

• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/isbn/1853126950
• Corporate Authors:

  WIT Press

  Ashurst Lodge
  Ashurst, Southampton  United Kingdom  SO40 7AA
• Authors:
  • Kaminski, W
  • Skrzypski, J
  • Strumillo
• Conference:
  • Fifth International Conference on Urban Transport and the Environment for the 21st Century
  • Location: Island of Rhodes, Greece
  • Date: 1999-9-8 to 1999-9-10
• Publication Date: 2000

Language

• English

Media Info

• Features: Figures; References; Tables;
• Pagination: p. 115-124
• Serial:
  • Publication of: WIT Press
  • Volume: 3
  • Publisher: WIT Press
  • ISSN: 1462-6101

Subject/Index Terms

• TRT Terms: Air pollution; Dimensional analysis; Forecasting; Neural networks; Particulates; Pollution control; Sulfur oxides; Suspended sediments; Urban areas; Weather conditions
• Subject Areas: Environment; Highways;

Filing Info

• Accession Number: 00789276
• Record Type: Publication
• ISBN: 1853126950
• Files: TRIS, ATRI
• Created Date: Mar 12 2000 12:00AM