Set-Membership ldentification of Hammerstein ModelsBased on l∞ Gain
Hiroaki FUKUSHIMA, Toshiharu SUGIE
In this paper, we propose a model set identification method for Hammerstein systems. This method gives a local model set near an equilibrium point by evaulating the l∞, gain of the model error for the given input level. The upper bound of l∞ gain can be obtained by a natural extension of an existing worst-case l1 identification method for linear systems, which is the main motivation to adopt l∞ gain as the uncertainty measure in this method. Also, this method gives less conservative model sets with more experimental data by using the noise set which consists of hard-bounded noises but takes account of a low correlation property of noise signals, simultaneously.