Identication of Wiener-Hammerstein systems by means of Support Vector Machines for Regression

1 AbstractIn this work we propose the use of Support Vector Ma-chines for Regression for the identi cation of Wiener-Hammerstein systems. The proposed methodology notonly leads to the de nition of an alternative tool for sys-tem identi cation, but gives also a better insight on thebehaviour of the Support Vector Machines approach.2 IntroductionA Wiener-Hammerstein system can be described, ingeneral terms, as the cascade of a rst linear dynamicsystem, a static nonlinear block, and a second lineardynamic system, as depicted in Fig.1.u p q