×

Identification of a class of nonlinear parametrically varying models. (English) Zbl 1021.93008

This paper proposes a class of nonlinear, possibly parameter-varying models suitable for system identification purposes. These models are given in the form of linear fractional transformation (LFT) where the “forward” part is represented by a conventional linear regression and the “feedback” part is given by a nonlinear dynamic map parametrized by a neural network. A parameter estimation based on the prediction error minimization has been developed. The algorithm exploits the seperation of the linear and nonlinear part.
The technique has been applied to a nonlinear identification problem arising in biomedical engineering, namely the one of determining accurate nonlinear models for knee joint dynamics in paraplectic patients.

MSC:

93B30 System identification
93C10 Nonlinear systems in control theory
92B20 Neural networks for/in biological studies, artificial life and related topics
92C10 Biomechanics
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Chen, International Journal of Control 50 pp 1873– (1989)
[2] and Identification, Adaptation and Learning?The Science of Learning Models From Data. Springer: Berlin, 1996. · doi:10.1007/978-3-662-03295-4
[3] Johansen, International Journal of Control 58 pp 1125– (1993)
[4] Previdi, Control Engineering Practice 7 pp 1489– (1999)
[5] Chen, International Journal of Control 49 pp 1013– (1989) · Zbl 0674.93009 · doi:10.1080/00207178908559683
[6] Sjoberg, Automatica 31 pp 1691– (1995)
[7] and Multiple Model Approaches to Modeling and Control, Taylor and Francis: London, 1997.
[8] Skeppstedt, International Journal of Control 55 pp 141– (1992)
[9] Johansen, Automatica 31 pp 321– (1995)
[10] and Identification of linear parameter-varying systems via LFT. In Proceedings of the 35th Conference on Decision and Control, Japan, 1996; 1545-1550.
[11] and Identification of linear parameter-varying Systems. In Proceeding of the 34th Conference on Decision and Control, 1995.
[12] and State space identification of MIMO linear parameter varying models. In Proceeding of the International Conference on Mathematical Theory of Networks and Systems, Padova, Italy, 1996.
[13] and Identification of a class of linear models with nonlinearly varying parameters. In Proceeding of the European Control Conference, Karlsruhe, Germany, 1999.
[14] and Identification of linear parameter varying models. In Proceeding of the Conference on Decision and Control, Phoenix, AZ, USA, 1999.
[15] and Robust and Optimal Control, Prentice-Hall: New Jersey, 1996. · Zbl 0999.49500
[16] Apkarian, Automatica 31 pp 1251– (1995)
[17] Apkarian, IEEE Transaction on Automatic Control 40 pp 853– (1995)
[18] Apkarian, IEEE Transactions on Control System Technology 6 pp 21– (1998)
[19] Hunt, Journal of Control 66 pp 619– (1997)
[20] Identification and Control using Local Linear Models, Ph.D. thesis. Politecnico di Milano, Italy, 1999.
[21] Kaminer, Automatica 31 pp 1185– (1995)
[22] Shamma, IEEE Transaction Automatic Control 35 pp 898– (1990)
[23] Narendra, IEEE Transaction on Neural Networks 2 pp 252– (1990)
[24] Cybenko, Mathematics in Control and Systems 2 pp 303– (1989)
[25] Practical Methods of Optimization, Wiley: New York, 1987. · Zbl 0905.65002
[26] Johansen, IEEE Control System Magazine 18 pp 48– (1998)
[27] Applications of control system technology in rehabilitation engineering involving functional electrical stimulation. In Invited Session at the European Control Conference 2001, Porto: Portugal, 2001.
[28] Riener, Bioelectrochemistry and Bioenergetics 43 pp 257– (1997)
[29] Riener, Philosophical Tnoransactions of the Royal Society of London Series B-Biological Sciences 354 pp 877– (1999)
[30] and Identification of a biomechanical system using neural networks. In Proceeding of the IFAC Workshop on Adaptation and Learning in Control and Signal Processing, Como, Italy, August 2001.
[31] and Linear modelling and controller design for the single limb movement of paraplegics using FES. In Preprints of the 4th IFAC Symposium on Modelling and Control in Biomedical Systems, Greifswald, Germany, March-April 2000; 7-12.
[32] and Application of a gain Scheduling Control Strategy to Artificial Stimulation of the Quadricepts Muscle. In European Control Conference, Karlsruhe, Germany, August-September 1999.
[33] and Identification of nonlinear models of artificial stimulation of the quadriceps muscle. In Proceeding of the IFAC Symposium on Modelling and Control in Biomedical Systems, Greifswald, Germany, March 2000; 83-89.
[34] Previdi, Control Engineering Practice 10 pp 91– (2002)
[35] Hunt, IEEE Transactions on Biomedical Engineering 45 pp 998– (1998)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.