Advertisement

Application of neural networks for classification of temperature distribution patterns

  • Henrik Saxén
  • Abhay Bulsari
  • Leif Karilainen
  • Kalevi Raipala
Conference paper

Abstract

Classification of spatially distributed measurements in industrial processes is an important but difficult task. Interpretation of the signals from horizontal probes in blast furnaces is a typical example. A neural network-based classification of such temperature measurements is described in this paper. Using temperature profiles first classified by the team, training and test sets were formed, and feedforward neural networks of different size were trained and evaluated on the material. The network size was found to clearly affect the quality of the classifications. By a judicious choice of the number of hidden nodes, a reliable neural classifier was obtained, which was incorporated in a supervision and control system for a Finnish blast furnace.

Keywords

Hide Layer Blast Furnace Hide Node Feedforward Neural Network Human Classification 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Poveromo, J.J., Milanoski, S.V. and Dwelly, M.J., Proc. 42th Ironmaking Conference, Atlanta, USA, pp. 491–501, ISS-AIME, 1983.Google Scholar
  2. 2.
    Nishio, H., Ariyama,T., Saito, H., Niwa, Y. and Yoshida, H., Proc. 41th Ironmaking Conference, Pittsburgh, USA, pp. 174–181, ISS-AIME, 1982.Google Scholar
  3. 3.
    Nagai, M., Saino, M., Tamura, S., Kaneka, K., Okabe, K., Fukutake, K. and Taguchi, S., Proc. 36th Ironmaking Conference, ISS-AIME, 1977.Google Scholar
  4. 4.
    Hirata, T., Yamamura, K., Morimoto, S. and Takada, H., Proc. 6th International Iron and Steel Congress, Nagoya, Japan, pp. 23–27, ISIJ 1990.Google Scholar
  5. 5.
    Otsuka, Y., Matsuda, K., Hanaoka, K., Kadoguchi, K. and Maki, T., Kobe Steel Engineering Reports 41 (4) 20–23, 1991.Google Scholar
  6. 6.
    Bulsari, A.B. and Saxén, H., Proc. Tenth International Conference on Systems Engineering (ICSE94), Coventry, England, pp. 157–163, 1994.Google Scholar
  7. 7.
    Karilainen, L., Saxén, H., Raipala, K. and Hinnelä, J., Proc. 53rd Ironmaking Conference, Chicago, USA, pp. 365–371, 1994.Google Scholar
  8. 8.
    Karilainen, L. and Saxén, H., Preprint 2nd IFAC Workshop on Computer Software Structures Integrating AI/KBS Systems in Process Control, Lund, Sweden, pp. 183–187, 1994.Google Scholar
  9. 9.
    Press, W.H., Flannery, B.P., Teukolsky, S.A. and Vetterling, W.T., Numerical Recipes, Cambridge University Press, Cambridge, USA, 1986.Google Scholar
  10. 10.
    Battiti, R., Neural Computation 4 (1992) 141–166.CrossRefGoogle Scholar
  11. 11.
    Bulsari, A.B. and Saxén, H., Neurocomputing 3 (1991) 125–133.MATHCrossRefGoogle Scholar
  12. 12.
    SL-GMS Reference Manual (Version 4.0), Sherrill-Lubinski Corporation, USA, 1991.Google Scholar

Copyright information

© Springer-Verlag/Wien 1995

Authors and Affiliations

  • Henrik Saxén
    • 1
  • Abhay Bulsari
    • 1
  • Leif Karilainen
    • 1
    • 2
  • Kalevi Raipala
    • 2
  1. 1.Dept. Chem. Eng.Åbo AkademiÅboFinland
  2. 2.FundiaKoverhar WorksLappvikFinland

Personalised recommendations