Abstract
This paper is concerned with identification problems in general, structured interconnected models. The elements to be identified are of two types: parametric components and non-parametric components. The non-parametric components do not have a natural parameterization that is known or suggested from an analytical understanding of the underlying process. These include static nonlinear maps and noise models.
We suggest a novel procedure for identifying such interconnected models. This procedure attempts to combine the best features of traditional parameter estimation and non-parametric system identification. The essential ingredient of our technique involves minimizing a cost function that captures the “smoothness” of the estimated nonlinear maps while respecting the available input-output data and the noise model. This technique avoids bias problems that arise when imposing artificial parameterizations on the nonlinearities. Computationally, our procedure reduces to iterative least squares problems together with Kalman smoothing.
Our procedure naturally suggests a scheme for control-oriented modeling. The problem here is to construct uncertainty models from data. The essential difficulty is to resolve residual signals into noise and unmodeled dynamics. We are able to do this using an iterative scheme. Here, we repeatedly adjust the noise component of the residual until its power-spectral density resembles that supplied in the noise model. Our computations again involve Kalman smoothing. This scheme is intimately related to our procedure for semi-parametric system identification.
Finally, we offer illustrative examples that reveal the promise of the techniques developed in this paper.
Supported in part by an NSF Graduate Fellowship, by NASA-Dryden under Contract NAG4-124, and by NSF under Grant ECS 95-09539.
This is a preview of subscription content, log in via an institution.
Preview
Unable to display preview. Download preview PDF.
References
S. A. Billings, S. Y. Fakhouri, “Identification of a Class of Non-linear Systems Using Correlation Analysis,” Proc. IEE, v.125, pp. 691–7, 1978.
M. Boutayeb, M. Darouach, H. Rafaralahy, G. Krzakala, “A New Technique for Identification of MISO Hammerstein Model,” Proc. American Control Conf., San Francisco, CA, v.2, pp. 1991–2, 1993.
C. Chen, S. Fassois, “Maximum likelihood identification of stochastic Wiener-Hammerstein-type non-linear systems,” Mech. Syst. Sig. Proc., vol. 6, no. 2, pp. 135–53, 1992.
J. E. Dennis, Jr. and R. B. Schnabel, Numerical methods for unconstrained optimization and nonlinear equations, Prentice-Hall, 1983.
R. Deutsch, Estimation Theory, Prentice-Hall, 1965.
D. K. De Vries, Identification of Model Uncertainty for Control Design, Ph.D. thesis, Delft University of Technology, Delft, the Netherlands.
D. K. De Vries and P. M. J. Van den Hof, “Quantification of uncertainty in transfer function estimation: a mixed probabilistic — worst-case approach,” Automatica, vol. 31, no. 4, pp. 543–557, 1995.
E. Wemhoff, “A Nonparametric Method for the Identification of a Static Nonlinearity in a Structured Nonlinear System,” M.S. Thesis, Department of Mechanical Engineering, University of California, Berkeley, August 1998.
A. H. Falkner, “Identification of the system comprising parallel Hammerstein branches,” Int. J. Syst. Sci., vol. 22, no. 11, pp. 2079–87, 1991.
B. Francis, A course in H ∞ control theory, Springer-Verlag, 1987.
P. E. Gill, W. Murray, and M. H. Wright, Practical optimization, Academic Press, 1981.
G. Golub and C. F. Van Loan, Matrix Computations, 2nd edition, The Johns Hopkins University Press, 1989.
G. C. Goodwin, M. Gevers and B. Ninness, “Quantifying the error in estimated transfer functions with application to model order selection,” IEEE Trans. Autom. Control, vol. 37, no. 7, pp. 913–928, 1992.
W. Greblicki, “Non-parametric orthogonal series identification of Hammerstein systems,” Int. J. Syst. Sci., vol. 20, no. 12, pp. 2355–67, 1989.
M. S. Grewal, Kalman filtering: theory and practice, Prentice-Hall, 1993.
H. Hjalmarsson, Aspects on Incomplete Modeling in System Identification, Ph.D. thesis, Electrical Engineering, Linköping University, Linköping, Sweden, 1993.
T. A. Johansen, “Identification of Non-linear Systems using Empirical Data and Prior Knowledge-An Optimization Approach,” Automatica, vol.32, no.3, pp. 337–56, 1997.
A. Juditsky, H. Hjalmarsson, A. Benveniste, B. Delyon, L. Ljung, J. Sjöberg, and Q. Zhang, “Nonlinear Black-box Models in System Identification: Mathematical Foundations,” Automatica, vol.31, no.12, pp. 1752–1750, 1995.
M. V. P. Krüger and K. Poolla, “Validation of uncertainty in the presence of noise,” Sel. Topics in Identification, Modeling and Control, vol. 11, Delft University Press, pp. 1–8, 1998.
A. Kryzak, “Identification of discrete Hammerstein systems by the Fourier series regression estimate,” Int. J. Syst. Sci., vol. 20, no. 9, pp. 1729–44, 1989.
L. Ljung, System Identification, Theory for the User, Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1987.
K. S. Narendra, P. G. Gallman, “An Iterative Method for the Identification of Nonlinear Systems Using the Hammerstein Model,” IEEE Trans. Autom. Control, vol. 11, no. 7, pp. 546–50, 1966.
A. V. Oppenheim and R. W. Schafer, Discrete-time signal processing, Prentice-Hall, 1989.
A. Poncet and G. S. Moschytz, “Selecting inputs and measuring nonlinearity in system identification,” Proc. Int. Workshop on Neural Networks for Identification, Control, Robotics, and Signal/Image Processing, Venice, pp. 2–10, 1996.
K. Poolla, P. Khargonekar, A. Tikku, J. Krause and K. Nagpal, “A time-domain approach to model validation,” IEEE Trans. Autom. Control, vol. 39, no.5, pp. 951–959, 1994.
M. Pawlak, “On the series expansion approach to the identification of Hammerstein systems,” IEEE Trans. Auto. Contr., vol. 36, no. 6, pp. 763–7, 1991.
R. Sen, P. Guhathakurta, “On the solution of nonlinear Hammerstein integral equation in L 2(0,1),” Trans. Soc. Comp. Simu., vol. 8, no. 2, pp. 75–86, 1991.
J. Sjöberg, Q. Zhang, L. Ljung, A. Benveniste, B. Delyon, P. Y. Glorennec, H. Hjalmarsson, and A. Juditsdy, “Nonlinear Black-box Modeling in System Identification: a Unified Overview,” Automatica, vol.31, no.12, pp. 1691–1724, 1995.
R. Smith, G. Dullerud, S. Rangan and K. Poolla, “Model validation for dynamically uncertain systems,” Mathematical Modelling of Systems, vol. 3, no. 1, pp. 43–58, 1997.
P. Stoica, “On the Convergence of an Iterative Algorithm Used for Hammerstein System Identification,” IEEE Trans. Autom. Control, vol. 26, no. 4, pp. 967–69, 1981.
G. Vandersteen and J. Schoukens, “Measurement and Identification of Nonlinear Systems consisting out of Linear Dynamic Blocks and One Static Nonlinearity,” IEEE Instrumentation and Measurement Technology Conference, vol.2, pp. 853–8, 1997.
G. Wolodkin, S. Rangan, and K. Poolla, “An LFT approach to parameter estimation,” Proc. American Control Conf., Albuquerque, NM, pp. 2088–2092, 1997.
Author information
Authors and Affiliations
Corresponding author
Editor information
Rights and permissions
Copyright information
© 1999 Springer-Verlag London Limited
About this paper
Cite this paper
Krüger, M., Wemhoff, E., Packard, A., Poolla, K. (1999). Semi-parametric methods for system identification. In: Garulli, A., Tesi, A. (eds) Robustness in identification and control. Lecture Notes in Control and Information Sciences, vol 245. Springer, London. https://doi.org/10.1007/BFb0109859
Download citation
DOI: https://doi.org/10.1007/BFb0109859
Published:
Publisher Name: Springer, London
Print ISBN: 978-1-85233-179-5
Online ISBN: 978-1-84628-538-7
eBook Packages: Springer Book Archive