RBFN identification of a solution copolymerization model

Author

Bomberger, John D. ; Seborg, Dale E. ; Ogunnaike, Babatunde A.

Author_Institution

Dept. of Chem. Eng., California Univ., Santa Barbara, CA, USA

Volume

5

fYear

1998

fDate

21-26 Jun 1998

Firstpage

3182

Abstract

Methods developed for radial basis function network (RBFN) identification are applied to a complex multiple-input, multiple-output (MIMO) simulation. For RBFN identification, stepwise regression analysis is used, together with model order determination using the method of false nearest neighbors and width parameter estimation using approximate gradient norms. Industrially practical input sequence design is also considered

Keywords

MIMO systems; autoregressive processes; chemical technology; feedforward neural nets; parameter estimation; polymerisation; process control; statistical analysis; MIMO simulation; RBFN; approximate gradient norms; complex multiple-input multiple-output simulation; false nearest neighbors method; input sequence design; model order determination; radial basis function network identification; solution copolymerization model; stepwise regression analysis; width parameter estimation; Chemical engineering; Continuous-stirred tank reactor; Input variables; MIMO; Nearest neighbor searches; Parameter estimation; Polymers; Radial basis function networks; Regression analysis; Solvents;

fLanguage

English

Publisher

ieee

Conference_Titel

American Control Conference, 1998. Proceedings of the 1998

Conference_Location

Philadelphia, PA

ISSN

0743-1619

Print_ISBN

0-7803-4530-4

Type

conf

DOI

10.1109/ACC.1998.688449

Filename

688449