Advertisement

Kohonen Network as a Classifier and Predictor for the Qualification of Metal-Oxide-Surfaces

  • Waltraud Kessler
  • Rudolf W. Kessler
Chapter
Part of the Advances in Industrial Control book series (AIC)

Abstract

The corrosion of metals and the paint adhesion on metals is a result of the superposition of complex reactions on the surface, which depend on surface oxide thickness, its porosity and chemical composition. By means of diffuse reflectance spectroscopy and evaluation of the spectra by a Kohonen self-organizing map, it is possible to predict the future corrosion behaviour of low carbon steel. Combining a Kohonen map and an interpolation method in the output layer allows to determine the layer thickness of conversion layers on aluminium from their interference spectra. This offers a fast, reliable and on-line applicable tool to calculate the thickness of transparent surface layers on aluminium or other metals even in the range below 100 nm.

Keywords

Radial Basis Function Diffuse Reflectance Spectrum Radial Basis Function Network Interference Spectrum Kohonen Network 
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.
    Kohonen, T.: Self-organization and associative memory. Third Edition, Springer Verlag, Berlin, Germany (1989)Google Scholar
  2. 2.
    Hecht-Nielsen, R.,: Counterpropagation Networks. Appl. Optics 26 (1987) 4979–4984CrossRefGoogle Scholar
  3. 3.
    Kessler, W., Göppert, J., Kessler, R. W.: Prediction of oxide layer thickness by a topology preserving interpolation method in a self-organizing map. Proceedings of the 7th International Conference on Systems Research, Informatics and Cybernetics, Baden-Baden (1994) 99–104Google Scholar
  4. 4.
    Göppert, J., Rosenstiel, W.,: Self-organizing maps vs. backpropagation: An experimental study. Proceedings of Design Methodologies for Microelectronics and Signal Processing, Giwice, Poland (1993) 153–162Google Scholar
  5. 5.
    Poggio, T., Girosi, F.: A theory of networks for approximation and learning. A.I. Memo No. 1140, MIT (1989)Google Scholar
  6. 6.
    Broomhead, D. S. and Lowe, D.: Multivariable functional interpolation and adaptive network. Complex Systems, 2 (1988) 321–355MathSciNetzbMATHGoogle Scholar
  7. 7.
    Ende, D., Kessler, W., Oelkrug, D., Fuchs, R.: Characterization of Chromate- phosphate conversion layers on Al-alloys by electrochemical impedance spectroscopy (EIS) and optical measurements. Electrochimica Acta, 38 (17) (1993) 2577–2580CrossRefGoogle Scholar
  8. 8.
    Hecht, E.: Optik. Addison-Wesley, Bonn (1989)Google Scholar
  9. 9.
    Vasicek, A.: Optics of thin Films. North-Holland Publishing Company (1960)zbMATHGoogle Scholar
  10. 10.
    Gauglitz, G., Brecht, A., Kraus, G., Nahm, W.: Chemical and biochemical sensors based on interferometry at thin (multi-)layers. Sensors and Actuators B, 11 (1993) 21–27CrossRefGoogle Scholar
  11. 11.
    Kessler, R. W., Böttcher, E., Füllemann, R., Oelkrug, D.: In situ characterisation of electrochemical formed oxide films on low carbon steel by diffuse reflectance spectroscopy. Fresenius Z. Anal. Chem 319 (1984) 695–700CrossRefGoogle Scholar
  12. 12.
    Kessler, R. W., Brögeler, M., Tubach, M., Degen, W. Zwick W.: Determination of the corrosion behaviour of car-body steel by optical methods. Werkstoffe und Korrosion 40 (1989) 539–544CrossRefGoogle Scholar
  13. 13.
    Kessler, R.W., Kessler, W., Quint, B., Kraus, M: C. Jochum (Ed.) Multivariate Analysis of unstructured diffuse reflectance spectra from car body steel surfaces. Software Development in Chemistry 8, Gesellschaft Deutscher Chemiker (1994) 91–98Google Scholar
  14. 14.
    Kessler, R.W, Degen, W., Zwick, W.: Optical on-line sensor to determine the corrosion behaviour of low carbon steel. Deutscher Verband für Materialforschung und -prüfung, Tagungsband Werkstoffprüfung, Bad Nauheim, Germany (1989) 329ffGoogle Scholar
  15. 15.
    Kessler, W. Kessler, R. W., Kraus, M. Kübier, R., Weinberger, K.: Improved prediction of the corrosion behaviour of car body steel using a Kohonen self-organizing map. Colloquium Digest IEE Colloquium on Advances on Neural Networks for Control and Systems, Berlin (1994)Google Scholar

Copyright information

© Springer-Verlag London Limited 1995

Authors and Affiliations

  • Waltraud Kessler
    • 1
  • Rudolf W. Kessler
    • 1
  1. 1.Institut für Angewandte ForschungFachhochschule ReutlingenReutlingenGermany

Personalised recommendations