Nonlinear Dynamic System Identification into Wiener Model using Subspace Identification and Support Vector Machine

In this article, a new block- oriented nonlinear identification method is proposed. This modeling method uses the Wiener model comprised of a linear dynamic block that is followed by a nonlinear static block. The linear block is described by the subspace identification algorithm whereas the nonlinear one is represented via the Least Squares- Support Vector Machine. The proposed method is tested with a practical nonlinear chemical plant named as CSTR. A dataset of the input-output signals gathered from the system is applied to show the superiority of the method.

Keywords:

Block-oriented identification , Wiener model , Subspace identification , Least Squares-Support Vector Machine , CSTR system