11 février 2021
Ce document est lié à :
Schriftenreihe des Instituts für Technische Mechanik, Karlsruher Institut für Technologie
Directory of Open Access Books, ID : 10670/1.wvo738
A nonparametric identification method for highly nonlinear systems is presented that is able to reconstruct the underlying nonlinearities without a priori knowledge of the describing nonlinear functions. The approach is based on nonlinear Kalman Filter algorithms using the well-known state augmentation technique that turns the filter into a dual state and parameter estimator, of which an extension towards nonparametric identification is proposed in the present work.