Title: Guaranteed frequency-domain identification of fractional order systems: application to a real system

Authors: Messaoud Amairi; Mohamed Aoun; Slaheddine Najar; Mohamed Naceur Abdelkrim

Addresses: National Engineering School of Gabes (ENIG), Research Unit Modelling, Analysis and Control of Systems (MACS), University of Gabes, Omar Ibn el Khattab Street, 6029 Gabes, Tunisia. ' National Engineering School of Gabes (ENIG), Research Unit Modelling, Analysis and Control of Systems (MACS), University of Gabes, Omar Ibn el Khattab Street, 6029 Gabes, Tunisia. ' National Engineering School of Gabes (ENIG), Research Unit Modelling, Analysis and Control of Systems (MACS), University of Gabes, Omar Ibn el Khattab Street, 6029 Gabes, Tunisia. ' National Engineering School of Gabes (ENIG), Research Unit Modelling, Analysis and Control of Systems (MACS), University of Gabes, Omar Ibn el Khattab Street, 6029 Gabes, Tunisia

Abstract: This paper presents a new guaranteed approach for frequency-domain identification of fractional order systems. Estimated parameters (coefficients and differential orders) are expressed as intervals. Then, an interval-based global optimisation algorithm is used to estimate the set of all feasible parameters. It combines the Hansen's algorithm with forward-backward contractor. The approach is applied to a numerical example as well as to a real electronic system.

Keywords: fractional differentiation; fractional systems; real interval analysis; global optimisation; frequency domain identification; constrained satisfaction problems; system identification; electronic systems.

DOI: 10.1504/IJMIC.2012.048637

International Journal of Modelling, Identification and Control, 2012 Vol.17 No.1, pp.32 - 42

Published online: 17 Dec 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article