Mathematical Foundations of Non-Classical Extensions of Similarity Theory

  • Stephan Rudolph
Conference paper
Part of the Iutam Bookseries book series (IUTAMBOOK, volume 10)


Similarity theory in form of the Pi-Theorem guarantees for any dimensionally homogeneous function the existence of a dimensionless similarity function of dimensionless parameters. Similarity theory is known for its successful applications in science and engineering. Classical applications of similarity theory in engineering mostly exploit the fact that two distinct objects or processes are said to be completely similar if their dimensionless parameters are identical. Similarity theory is used extensively in mathematics for the invariance analysis of differential equations and the derivation of exact and/or approximate solutions. In this work, an extension of the classical applications of similarity theory to artificial intelligence, notably in the fields of case-based and rule-based reasoning, neural networks, data mining, pattern recognition and sound classification, is presented. It is shown that the validity of the extension can be guaranteed by a so-called embedding theorem based on the property of dimensional homogeneity.


Similarity theory Pi-theorem embedding theorem artificial intelligence 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Buckingham, E. “On physically similar systems: Illustrations of the use of dimensional equations”, Physical Review, 4, pp. 345–376, 1914.CrossRefGoogle Scholar
  2. 2.
    Bridgman, P. Dimensional Analysis. New Haven, Yale University Press, 1922.Google Scholar
  3. 3.
    Vaschy, A. “Sur les lois de similitude en Physique”, Annales Télégraphiques (3e série), 19, pp. 25–28, 1892. (French)Google Scholar
  4. 4.
    Federman, A. “On some general methods of integration of partial differential equations of the first order”, Izvestiya St. Petersburgh Polytechn. Inst, 16, pp. 97–155, 1911. (in Russian)Google Scholar
  5. 5.
    Riabouchinsky, D.P. “Méthode des variables de dimension zéro, et son application en aérodynamique”, L’Aérophile, 1, Septembre, 407–408, 1911. (in French)Google Scholar
  6. 6.
    Görtler, H. “Zur Geschichte des Pi-Theorems”, ZAMM, 55, pp. 3–8, 1975. (in German)zbMATHCrossRefGoogle Scholar
  7. 7.
    Görtler, H. Dimensionsanalyse, Berlin, Springer, 1975. (in German)zbMATHGoogle Scholar
  8. 8.
    Rudolph, S. Übertragung von Ähnlichkeitsbegriffen, Habilitationsschrift, Fakultät für Luft- und Raumfahrttechnik und Geodäsie, Universität Stuttgart, 2002. (in German)Google Scholar
  9. 9.
    Langhaar, H. Dimensional Analysis and Theory of Models, New York, John Wiley, 1951.zbMATHGoogle Scholar
  10. 10.
    Baker, W., Westine, P. and Dodge, F. Similarity Methods in Engineering Dynamics. Theory and Practice of Scale Modeling. Amsterdam, Elsevier, 1991.Google Scholar
  11. 11.
    Bluman, G. and Cole, J. Similarity Methods for Differential Equations, New York, Springer, 1974.zbMATHGoogle Scholar
  12. 12.
    Bluman, G. and Kumei, S. Symmetries and Differential Equations. New York, Springer, 1989.zbMATHGoogle Scholar
  13. 13.
    Barenblatt, G. Scaling, Self-Similarity, and Intermediate Asymptotics, Cambridge, Cambridge University, Press, 1996.zbMATHGoogle Scholar
  14. 14.
    Kline, S. Similitude and Approximation Theory. Berlin, Springer, 1986.Google Scholar
  15. 15.
    Huntley, H.Dimensional Analysis. London, MacDonald, 1952.zbMATHGoogle Scholar
  16. 16.
    Duncan, W. Physical Similarity and Dimensional Analysis. London, Arnold, 1953.zbMATHGoogle Scholar
  17. 17.
    Mittelstrass, J. (Hrsg.) Enzyklopädie Philosophie und Wissenschaftstheorie, 4 Bände, Stuttgart, Metzler, 1995. (Satz vom ausgeschlossenen ⇑Widerspruch, in German)Google Scholar
  18. 18.
    Kolodner, J. Case-Based Reasoning, San Mateo, Morgan Kaufmann, 1993.Google Scholar
  19. 19.
    Rudolph, S. “On the foundations and applications of similarity theory to case-based reasoning”, Proceedings of the 12th International Conference for Applications of Artificial Intelligence in Engineering (AIENG′97), Capri (Naples), Italy, July 7–9, 1997.Google Scholar
  20. 20.
    Hertkorn, P. and Rudolph, S. “Dimensional analysis in case-based reasoning”, Proceedings International Workshop on Similarity Methods, University of Stuttgart, Germany, November 26–27, pp. 163–178, 1998.Google Scholar
  21. 21.
    Sanchez-Sinencio, E. and Lau, C. (eds), Artificial Neural Networks, New York, IEEE Press, 1992.Google Scholar
  22. 22.
    Hornik, K., Stinchcombe, M. and White, H. “Multilayer feed-forward networks are universal approximators”, Neural Networks, 2, 5, pp. 359–366, 1989.CrossRefGoogle Scholar
  23. 23.
    Rudolph, S. “On topology, size and generalization in non-linear feed-forward neural networks”, Neurocomputing, 16, 1, pp. 1–22, July 1997.Google Scholar
  24. 24.
    Rudolph, S. “On a data-driven model identification technique using artificial neural networks”, Proceedings of EUROMECH 373 Colloquium on Modelling and Control of Adaptive Mechanical Structures, Magdeburg, Germany, March 11–13, Fortschritt-Berichte VDI, Reihe 11, Nummer 268, pp. 341–350, 1998.Google Scholar
  25. 25.
    Adriaans, P and Zantinge, D. Data Mining. Harlow, Addison-Wesley, 1996.Google Scholar
  26. 26.
    Fayyad, U., et al. (eds.) Advances in Knowledge Discovery and Data Mining. Menlo Park, CA: AAAI/MIT Press, 1996.Google Scholar
  27. 27.
    Hertkorn, P. Knowledge Discovery in Databases auf der Grundlage dimensionshomogener Funktionen, PhD Thesis, Fakultät Luft- und Raumfahrttechnik und Geodäsie, Universität Stuttgart, Stuttgart, 2004. (in German)Google Scholar
  28. 28.
    Rudolph, S. and Hertkorn, P. “Data mining in scientific data”. In: Data Mining for Design and Manufacturing: Methods and Applications, Braha, D. (ed), pp. 61–85, Dordrecht, Kluwer, 2002.Google Scholar
  29. 29.
    Brückner, S. and Rudolph, S. “Knowledge discovery in scientific data using hierarchical modeling in dimensional analysis”. Proceedings SPIE Aerosense 2001 Conference On Data Mining and Knowledge Discovery III, Orlando, FL, April 16–20, 2001.Google Scholar
  30. 30.
    Rudolph, S. “Knowledge discovery in scientific data”, Proceedings SPIE Aerosense Conf. On Data Mining and Knowledge Discovery II, Orlando, FL, ,cm.April 24–28, 2000.Google Scholar
  31. 31.
    Melan, A. N-dimensionale Merkmalsgewinnung durch vektorielle Dimensionen am Beispiel von Farbbildern und weiteren Anwendungen, PhD Thesis, Fakultät Luft- und Raumfahrttechnik und Geodäsie, Universität Stuttgart, Stuttgart, 2004. (in German)Google Scholar
  32. 32.
    Till, M. and Rudolph, S. Optimized time-frequency distributions for signal classification with feed-forward neural networks, Proceedings SPIE Aerosense Conference On Applications and Science of Computational Intelligence III, Orlando, FL, April 24–28, 2000.Google Scholar
  33. 33.
    Till, M. Geräuschklassifikation mittels Übertragung physikalischerÄhnlichkeit, PhD Thesis, Fakultät Luft- und Raumfahrttechnik und Geodäsie, Universität Stuttgart, Stuttgart, 2007. (in German)Google Scholar

Copyright information

© Springer Science+Business Media B.V. 2009

Authors and Affiliations

  • Stephan Rudolph
    • 1
  1. 1.Priv.-Doz. Dr.-Ing., Similarity Mechanics Group HeadInstitute for Statics and Dynamics of Aerospace Structures University of StuttgartGermany

Personalised recommendations