A neural network model for blast furnace wall temperature pattern classification

  • Henrik Saxén
  • Leif Lassus
Conference paper


A model for classification, visualization and interpretation of temperature distributions from the wall of the ironmaking blast furnace is presented. The model classifies the patterns using a self-organizing map and depicts the evolution of the distributions on the feature map, which is used as an operation diagram. The model has been implemented at the blast furnaces of a Finnish steelmaking company to improve the alertness of the operators and to help them to take appropriate control actions. The generic features of the models make it possible to apply the proposed classification method to different furnaces with only minor overhead for model tuning. Use of the classifications in operation diagrams is finally discussed.


Blast Furnace Wall Temperature Blast Furnace Operation Model Tuning Operation Diagram 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    Biswas, A.K.: Principles of Blast Furnace Ironmaking. Brisbane: Cootha Publishing House 1981.Google Scholar
  2. [2]
    Omori, Y., et al.: Blast Furnace Phenomena and Modeling. London: Elsevier Applied Science 1987.Google Scholar
  3. [3]
    Hirata T., et al.:Blast Furnace Operation System using Neural Networks and Knowledge-Base. Proc. 6th International Iron and Steel Congress, Nagoya, Japan, pp. 23–27, 1990.Google Scholar
  4. [4]
    Takada H., et al.: A Design Environment for Neural Networks and Its Applications. Control Engineering Practice, Vol. 2, pp. 123–128, 1994.CrossRefGoogle Scholar
  5. [5]
    Bulsari, A.B., Saxén H.: Classification of blast furnace probe temperatures using neural networks. Steel Research, Vol. 66, pp. 231–236, 1995.Google Scholar
  6. [6]
    Otsuka Y., et al.: Application of Neural Network Systems to Pattern Recognition of Blast Furnace Operation Data. Kobelco Technology Review, Vol. 15, pp. 12–16, 1992.Google Scholar
  7. [7]
    Kohonen, T.: Self-Organization and Associative Memory. Berlin: Springer-Verlag 1989.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Wien 1999

Authors and Affiliations

  • Henrik Saxén
    • 1
  • Leif Lassus
    • 2
  1. 1.Heat Engineering LaboratoryÅbo Akademi UniversityÅboFinland
  2. 2.TT Tieto Oy, TelecomHelsinkiFinland

Personalised recommendations