Authors:
Przemysław Śliwiński
and
Zygmunt Hasiewicz
Affiliation:
Wrocław University of Technology, Poland
Keyword(s):
Hammerstein system, Non-parametric identification, Orthogonal expansions, Regression estimation, Computational algorithm.
Related
Ontology
Subjects/Areas/Topics:
Informatics in Control, Automation and Robotics
;
Nonlinear Signals and Systems
;
Signal Processing, Sensors, Systems Modeling and Control
;
System Identification
Abstract:
In the paper a fast computational routines for identification algorithms for recovering nonlinearities in Hammerstein systems based on orthogonal series expansions of functions are proposed. It is ascertained that both, convergence conditions and convergence rates of the computational algorithms are the same as their much less computationaly attractive 'theoretic' counterparts. The generic computational algorithm is derived and illustrated by three examples based on standard orthogonal series on interval, viz. Fourier, Legendre, and Haar systems. The exemplary algorithms are presented in a detailed, ready-to-implement, form and examined by means of computer simulations.