The identification of nonlinear dynamic systems around operating points using neural networks

Author

Pienaar, J.D. ; Bodenstein, C.P.

Author_Institution

Sch. of Electr. & Electron. Eng., Potchefstroom Univ. for CHE, South Africa

Volume

1

fYear

1998

fDate

7-10 Jul 1998

Firstpage

105

Abstract

This paper discusses a method of modeling the dynamic relationships of a nonlinear time-invariant or slowly varying system, typical of those found in the petrochemical industry. A number of setpoints are identified, and linear models are constructed around these points. Every linear model is then used to construct a NARMAX model (nonlinear autoregressive moving average model with autogeneous inputs). Finally a chemical process is modeled to illustrate the concepts illustrated in this paper

Keywords

autoregressive moving average processes; chemical industry; identification; neural nets; nonlinear dynamical systems; NARMAX model; autogeneous inputs; chemical process modeling; linear models; neural networks; nonlinear autoregressive moving average model; nonlinear dynamic systems identification; nonlinear time-invariant system; operating points; slowly varying system; Africa; Autoregressive processes; Chemical industry; Control system synthesis; Electrical equipment industry; Neural networks; Nonlinear dynamical systems; Petrochemicals; Predictive models; System identification;

fLanguage

English

Publisher

ieee

Conference_Titel

Industrial Electronics, 1998. Proceedings. ISIE '98. IEEE International Symposium on

Conference_Location

Pretoria

Print_ISBN

0-7803-4756-0

Type

conf

DOI

10.1109/ISIE.1998.707757

Filename

707757