From linear to nonlinear identification: One step at a time

We propose a model structure based on High Level Canonical Piecewise Linear (HL CPWL) functions and a corresponding Nonlinear Output Error (NOE) identification algorithm. We explore the approximation capabilities of this novel structure together with its generalization and stability properties. Starting from a linear Output Error (OE) approximation, this model family yields an identification algorithm such that the order of the model can be easily increased during the identification process, retaining the previously achieved approximation. The parameters of the HL CPWL functions are learned using a simple algorithm that guarantees BIBO stability of the model. Applying our method is very straightforward and constitutes a principled approach for transitioning from linear to nonlinear model structures. Furthermore, it offers the possibility of an efficient hardware implementation in VLSI. A few examples are provided in this article in order to demonstrate the potential of our approach.

[1]  O. Nelles.  Nonlinear System Identification , 2001 .

[2]  Pedro Julian,et al.  Synthesis of multiport resistors with piecewise-linear characteristics: a mixed-signal architecture: Research Articles , 2005 .

[3]  Jakob Roll.  Local and Piecewise Affine Approaches to System Identification , 2003 .

[4]  Johan A. K. Suykens,et al.  Nonlinear modeling : advanced black-box techniques , 1998 .

[5]  Jonas Sjöberg,et al.  Neural Nets and Related Model Structures for Nonlinear System Identification , 1998 .

[6]  Marco Storace,et al.  Piecewise-Linear Identification of Nonlinear Dynamical Systems in View of Their Circuit Implementations , 2007, IEEE Transactions on Circuits and Systems I: Regular Papers.

[7]  Khashayar Khorasani,et al.  Adaptive time delay neural network structures for nonlinear system identification , 2002, Neurocomputing.

[8]  Liliana Raquel Castro,et al.  Rational wavelets in Wiener-like modeling , 2002 .

[9]  Stephen A. Billings,et al.  International Journal of Control , 2004 .

[10]  M. Korenberg,et al.  Orthogonal approaches to time-series analysis and system identification , 1991, IEEE Signal Processing Magazine.

[11]  Alfredo C. Desages,et al.  Wiener modelling using canonical pieciewise linear functions , 1999 .

[12]  P. Julián,et al.  High-level canonical piecewise linear representation using a simplicial partition , 1999 .

[13]  Pedro Julián,et al.  Synthesis of multiport resistors with piecewise-linear characteristics: a mixed-signal architecture , 2005, Int. J. Circuit Theory Appl..

[14]  Stephen A. Billings,et al.  Properties of neural networks with applications to modelling non-linear dynamical systems , 1992 .

[15]  Roderick Murray-Smith,et al.  Multiple Model Approaches to Modelling and Control , 1997 .

[16]  Eduardo Sontag.  Nonlinear regulation: The piecewise linear approach , 1981 .

[17]  Lennart Ljung,et al.  Nonlinear black-box modeling in system identification: a unified overview , 1995, Autom..

[18]  T Poggio,et al.  Regularization Algorithms for Learning That Are Equivalent to Multilayer Networks , 1990, Science.

[19]  Heinz Unbehauen,et al.  Structure identification of nonlinear dynamic systems - A survey on input/output approaches , 1990, Autom..

[20]  Sheng Chen,et al.  Representations of non-linear systems: the NARMAX model , 1989 .

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

[22]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[23]  C. Fantuzzi,et al.  Identification of piecewise affine models in noisy environment , 2002 .

[24]  P. Julian,et al.  VLSI microprocessor architecture for a simplicial PWL function evaluation core , 2008, 2008 Argentine School of Micro-Nanoelectronics, Technology and Applications.