Information-Dependent Switching of Identification Criteria in a Genetic Programming System for System Identification

  • Thomas Buchsbaum
  • Siegfried Vössner
Part of the Lecture Notes in Computer Science book series (LNCS, volume 3905)


Genetic Programming (GP) can be used to identify the nonlinear differential equations of dynamical systems. If, however, the fitness function is chosen in a classical way, the optimization will not work very well. In this article, we explain the reasons for the failure of the GP approach and present a solution strategy for improving performance. Using more than one identification criterion (fitness function) and switching based on the information content of the data enable standard GP algorithms to find better solutions in shorter times. A computational example illustrates that identification criteria switching has a bigger influence on the results than the choice of the GP parameters has.


Genetic Programming System Criterion Switching Genetic Programming Algorithm Genetic Programming Approach Standard Genetic Programming 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Babovic, V., Keijzer, M.: Evolutionary algorithms approach to induction of differential equations. In: Proceedings of the Fourth International Conference on Hydroinformatics, Iowa City, USA (2000)Google Scholar
  2. 2.
    Burgess, G.: Finding Approximate Analytic Solutions To Differential Equations Using Genetic Programming, Salisbury, Australia, vol. 5108 (1999)Google Scholar
  3. 3.
    Cao, H., Kang, L., Chen, Y.: Evolutionary Modeling of Ordinary Differential Equations for Dynamic Systems. In: Proceedings of the Genetic and Evolutionary Computation Conference, Orlando, Florida, USA, vol. 2, pp. 959–965. Morgan Kaufmann, San Francisco (1999)Google Scholar
  4. 4.
    Gagne, C., Parizeau, M.: Open BEAGLE: A New C++ Evolutionary Computation Framework (2002)Google Scholar
  5. 5.
    Gagne, C., Parizeau, M.: Open BEAGLE: A New Versatile C++ Framework for Evolutionary Computations (2002)Google Scholar
  6. 6.
    Gray, G.J., Murray-Smith, D.J., Li, Y., Sharman, K.C., Weinbrenner, T.: Nonlinear model structure identification using genetic programming. Control Engineering Practice 6, 1341–1352 (1998)CrossRefGoogle Scholar
  7. 7.
    Haykin, S.: Neural Networks - A Comprehensive Foundation, 2nd edn., pp. 215–217. Prentice Hall International Inc., Englewood Cliffs (1999)MATHGoogle Scholar
  8. 8.
    Juang, J.N.: Applied system identification. Prentice-Hall, Inc., Upper Saddle River (1994)MATHGoogle Scholar
  9. 9.
    Keijzer, M., Babovic, V.: Dimensionally Aware Genetic Programming. In: Proceedings of the Genetic and Evolutionary Computation Conference, Orlando, Florida, USA, vol. 2, pp. 1069–1076. Morgan Kaufmann, San Francisco (1999)Google Scholar
  10. 10.
    Koza, J.R.: Genetic programming: On the programming of computers by natural selection. MIT Press, Cambridge (1992)MATHGoogle Scholar
  11. 11.
    Ljung, L.: System Identification: Theory for the User (Second Ed.). Prentice-Hall, Inc, Upper Saddle River (1998)CrossRefGoogle Scholar
  12. 12.
    Nelles, O.: Nonlinear System Identification, pp. 182–184. Springer, Berlin (2001)CrossRefMATHGoogle Scholar
  13. 13.
    Rodriguez-Vazquez, K., Fonseca, C.M., Fleming, P.J.: Multiobjective Genetic Programming: A Nonlinear System Identification Application. In: Late Breaking Papers at the 1997 Genetic Programming Conference, CA, USA, pp. 207–212. Stanford Bookstore, Stanford University (1997)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Thomas Buchsbaum
    • 1
  • Siegfried Vössner
    • 1
  1. 1.Department of Engineering and Business InformaticsGraz University of TechnologyGrazAustria

Personalised recommendations