A method for constructing NARMAX models of a class of nonlinear systems

Author

Seidel, Detlev K. ; Davies, Patricia

Author_Institution

Sch. of Mech. Eng., Purdue Univ., West Lafayette, IN, USA

fYear

1988

fDate

11-14 Apr 1988

Firstpage

2272

Abstract

A method of constructing nonlinear autoregressive moving average (NARMAX) models is described. The systems considered are linear subsystems with a polynomial feedback. The terms of the Volterra series of the response are generated sequentially, performing linear operations on nonlinear combinations of previous terms in the series. The z-transform of the linear operator is used to formulate a straightforward mapping formula from the coefficients of the nonlinear differential equation to the parameters in the NARMAX model. Numerical results from the application of the algorithm to a nonlinear oscillator show the effectiveness of the technique

Keywords

feedback; nonlinear systems; parameter estimation; series (mathematics); statistical analysis; NARMAX models; Volterra series; linear subsystems; mapping formula; nonlinear autoregressive moving average; nonlinear differential equation; nonlinear oscillator; nonlinear systems; polynomial feedback; z-transform; Autoregressive processes; Differential equations; Feedback; Laboratories; Linear systems; Mechanical engineering; Nonlinear equations; Nonlinear systems; Oscillators; Polynomials;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1988. ICASSP-88., 1988 International Conference on

Conference_Location

New York, NY

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1988.197090

Filename

197090