Quasi-parametric Recovery of Hammerstein System Nonlinearity by Smart Model Selection

  • Zygmunt Hasiewicz
  • Grzegorz Mzyk
  • Przemysław Śliwiński
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6114)


In the paper we recover a Hammerstein system nonlinearity. Hammerstein systems, incorporating nonlinearity and dynamics, play an important role in various applications, and effective algorithms determining their characteristics are not only of theoretical but also of practical interest. The proposed algorithm is quasi-parametric, that is, there are several parametric model candidates and we assume that the target nonlinearity belongs to the one of the classes represented by the models. The algorithm has two stages. In the first, the neural network is used to recursively filter (estimate) the nonlinearity from the noisy measurements. The network serves as a teacher/trainer for the model candidates, and the appropriate model is selected in a simple tournament-like routine. The main advantage of the algorithm over a traditional one stage approach (in which models are determined directly from measurements), is its small computational overhead (as computational complexity and memory occupation are both greatly reduced).


system identification structure detection Hammerstein system wavelet neural network 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Hasiewicz, Z., Mzyk, G.: Combined parametric-nonparametric identification of Hammerstein systems. IEEE Transactions on Automatic Control 49(8), 1370–1375 (2004)CrossRefMathSciNetGoogle Scholar
  2. 2.
    Hasiewicz, Z., Mzyk, G.: Hammerstein system identification by nonparametric instrumental variables. International Journal of Control 82(3), 440–455 (2009)zbMATHCrossRefMathSciNetGoogle Scholar
  3. 3.
    Śliwiński, P., Rozenblit, J., Marcellin, M.W., Klempous, R.: Wavelet amendment of polynomial models in nonlinear system identification. IEEE Transactions on Automatic Control 54(4), 820–825 (2009)CrossRefGoogle Scholar
  4. 4.
    Krzyzak, A., Linder, T., Lugosi, C.: Nonparametric estimation and classification using radial basis function nets and empirical risk minimization. IEEE Transactions on Neural Networks 7(2), 475–487 (1996)CrossRefGoogle Scholar
  5. 5.
    Kamiński, W., Strumiłło, P.: Kernel orthonormalization in radial basis function neural networks. IEEE Transactions on Neural Networks 8(5), 1177–1183 (1997)CrossRefGoogle Scholar
  6. 6.
    Krzyzak, A., Linder, T.: Radial basis function networks and complexity regularization in function learning. IEEE Transactions on Neural Networks 9(2), 247–256 (1998)CrossRefGoogle Scholar
  7. 7.
    Györfi, L., Kohler, M., Krzyżak, A., Walk, H.: A Distribution-Free Theory of Nonparametric Regression. Springer, New York (2002)zbMATHCrossRefGoogle Scholar
  8. 8.
    Buhmann, M.D.: Radial Basis Functions: Theory and Implementations. Cambridge University Press, Cambridge (2003)zbMATHCrossRefGoogle Scholar
  9. 9.
    Ferrari, S., Maggioni, M., Borghese, N.A.: Multiscale approximation with hierarchical radial basis functions networks. IEEE Transaction one Neural Networks 15(1) (2004)Google Scholar
  10. 10.
    Zhang, Q., Benveniste, A.: Wavelet networks. IEEE Transactions on Neural Networks 3, 889–898 (1992)CrossRefGoogle Scholar
  11. 11.
    Zhang, J., Walter, G., Miao, Y., Lee, W.N.: Wavelet neural networks for function learning. IEEE Transactions on Signal Processing 43, 1485–1497 (1995)CrossRefGoogle Scholar
  12. 12.
    Hasiewicz, Z.: Modular neural networks for non-linearity recovering by the Haar approximation. Neural Networks 13, 1107–1133 (2000)CrossRefGoogle Scholar
  13. 13.
    Greblicki, W., Pawlak, M.: Identification of discrete Hammerstein system using kernel regression estimates. IEEE Transactions on Automatic Control 31, 74–77 (1986)zbMATHCrossRefGoogle Scholar
  14. 14.
    Rutkowski, L.: Generalized regression neural networks in time-varying environment. IEEE Transactions on Neural Networks 15(3), 576–596 (2004)CrossRefMathSciNetGoogle Scholar
  15. 15.
    Śliwiński, P., Hasiewicz, Z.: Recursive wavelet estimation of Hammerstein systems nonlinearity. International Journal of Control (2010) (in preaparation)Google Scholar
  16. 16.
    Greblicki, W., Mzyk, G.: Semiparametric approach to Hammerstein system identification. In: Proceedings of the 15th IFAC Symposium on System Identification, Saint-Malo, France, pp. 1680–1685 (2009)Google Scholar
  17. 17.
    Rutkowski, L.: Adaptive probabilistic neural networks for pattern classification in time-varying environment. IEEE Transactions on Neural Networks 15(4), 811–827 (2004)CrossRefMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Zygmunt Hasiewicz
    • 1
  • Grzegorz Mzyk
    • 1
  • Przemysław Śliwiński
    • 1
  1. 1.Institute of Computer Engineering, Control and RoboticsWrocław University of TechnologyWrocławPoland

Personalised recommendations