On the interpretation and identification of dynamic Takagi-Sugeno fuzzy models

Dynamic Takagi-Sugeno fuzzy models are not always easy to interpret, in particular when they are identified from experimental data. It is shown that there exists a close relationship between dynamic Takagi-Sugeno fuzzy models and dynamic linearization when using affine local model structures, which suggests that a solution to the multiobjective identification problem exists. However, it is also shown that the affine local model structure is a highly sensitive parametrization when applied in transient operating regimes. Due to the multiobjective nature of the identification problem studied here, special considerations must be made during model structure selection, experiment design, and identification in order to meet both objectives. Some guidelines for experiment design are suggested and some robust nonlinear identification algorithms are studied. These include constrained and regularized identification and locally weighted identification. Their usefulness in the present context is illustrated by examples.

[1]  Tor Arne Johansen,et al.  A software environment for gain scheduled controller design , 1998 .

[2]  Robert Babuska.  Fuzzy Modelling and Identification , 1997 .

[3]  Tor Arne Johansen,et al.  Off-equilibrium linearisation and design of gain-scheduled control with application to vehicle speed control , 1998 .

[4]  Thomas Kailath,et al.  Linear Systems , 1980 .

[5]  John Yen,et al.  Improving the interpretability of TSK fuzzy models by combining global learning and local learning , 1998, IEEE Trans. Fuzzy Syst..

[6]  Winfried Stefan Lohmiller,et al.  Contraction analysis of nonlinear systems , 1999 .

[7]  Jean-Jacques E. Slotine,et al.  On Contraction Analysis for Non-linear Systems , 1998, Autom..

[8]  Tor Arne Johansen,et al.  Fuzzy model based control: stability, robustness, and performance issues , 1994, IEEE Trans. Fuzzy Syst..

[9]  Tor Arne Johansen,et al.  Local learning in local model networks , 1995 .

[10]  L. Ljung,et al.  Overtraining, Regularization, and Searching for Minimum in Neural Networks , 1992 .

[11]  L. Ljung,et al.  Overtraining, regularization and searching for a minimum, with application to neural networks , 1995 .

[12]  Dimiter Driankov,et al.  A Takagi-Sugeno fuzzy gain-scheduler , 1996, Proceedings of IEEE 5th International Fuzzy Systems.

[13]  R. Rovatti,et al.  On the approximation capabilities of the homogeneous Takagi-Sugeno model , 1996, Proceedings of IEEE 5th International Fuzzy Systems.

[14]  Wilson J. Rugh,et al.  Gain scheduling dynamic linear controllers for a nonlinear plant , 1995, Autom..

[15]  Jianqing Fan,et al.  Local polynomial modelling and its applications , 1994 .

[16]  Robert Shorten,et al.  On the interpretation of local models in blended multiple model structures. , 1999 .

[17]  W. Leithead,et al.  Analytic framework for blended multiple model systems using linear local models , 1999 .

[18]  Li-Xin Wang.  Stable adaptive fuzzy control of nonlinear systems , 1993, IEEE Trans. Fuzzy Syst..

[19]  Lars Kai Hansen,et al.  Generalization performance of regularized neural network models , 1994, Proceedings of IEEE Workshop on Neural Networks for Signal Processing.

[20]  L X Wang,et al.  Fuzzy basis functions, universal approximation, and orthogonal least-squares learning , 1992, IEEE Trans. Neural Networks.

[21]  Mauro Birattari,et al.  Lazy learning for modeling and control design , 1997 .

[22]  Tor Arne Johansen,et al.  On Tikhonov regularization, bias and variance in nonlinear system identification , 1997, Autom..

[23]  Tor Arne Johansen,et al.  Design and analysis of gain-scheduled control using local controller networks , 1997 .

[24]  Kazuo Tanaka,et al.  An approach to fuzzy control of nonlinear systems: stability and design issues , 1996, IEEE Trans. Fuzzy Syst..

[25]  M. Birattari,et al.  Lazy learning for local modelling and control design , 1999 .

[26]  Lennart Ljung,et al.  System Identification: Theory for the User , 1987 .

[27]  Wilson J. Rugh,et al.  Analytical Framework for Gain Scheduling , 1990, 1990 American Control Conference.

[28]  S. Billings,et al.  Piecewise linear identification of non-linear systems , 1987 .

[29]  Rainer Palm,et al.  Advances in Fuzzy Control , 1998 .

[30]  T. A. Johansen,et al.  Robust identification of Takagi-Sugeno-Kang fuzzy models using regularization , 1996, Proceedings of IEEE 5th International Fuzzy Systems.

[31]  R. Rovatti,et al.  Takagi-Sugeno models as approximators in Sobolev norms: the SISO case , 1996, Proceedings of IEEE 5th International Fuzzy Systems.

[32]  D. Jordan,et al.  Nonlinear Ordinary Differential Equations: An Introduction for Scientists and Engineers , 1979 .

[33]  T. Johansen,et al.  Constructing NARMAX models using ARMAX models , 1993 .

[34]  Tor Arne Johansen,et al.  Constrained and Regularized System Identification , 1997 .

[35]  T. McKelveyDepartment.  On the Use of Regularization in System Identification , 1992 .

[36]  Michio Sugeno,et al.  Fuzzy identification of systems and its applications to modeling and control , 1985, IEEE Transactions on Systems, Man, and Cybernetics.

[37]  E. Polak.  Introduction to linear and nonlinear programming , 1973 .

[38]  Stephen Yurkovich,et al.  Fuzzy Control , 1997 .

[39]  Fredrik Gustafsson,et al.  Just in time models for dynamical systems , 1996, Proceedings of 35th IEEE Conference on Decision and Control.

[40]  H. Müller,et al.  Local Polynomial Modeling and Its Applications , 1998 .

[41]  Tor Arne Johansen,et al.  Constrained and regularized system identification , 1997 .

[42]  T. Johansen,et al.  On transient dynamics, off-equilibrium behaviour and identification in blended multiple model structures , 1999, 1999 European Control Conference (ECC).