Title: A multiple projection approach for constrained system identification

Authors: M.F. Hassan; M. Zribi; H.M.K. Alazemi; M. Tawfik

Addresses: Electrical Engineering Department, College of Engineering and Petroleum, Kuwait University, P.O. Box 5969, Safat 13060, Kuwait. ' Electrical Engineering Department, College of Engineering and Petroleum, Kuwait University, P.O. Box 5969, Safat 13060, Kuwait. ' Electrical Engineering Department, College of Engineering and Petroleum, Kuwait University, P.O. Box 5969, Safat 13060, Kuwait. ' Electrical Engineering Department, College of Engineering and Petroleum, Kuwait University, P.O. Box 5969, Safat 13060, Kuwait

Abstract: This paper deals with the constrained system identification problem of linear discrete time dynamical systems. It is assumed that the parameters of the system are constrained due to physical limitations. Using a multiple projection approach, a minimum variance estimator and its associated recursive version are developed to estimate the constrained parameters of the system. The most important feature of the developed technique is that it gives estimates of the system parameters once the first set of measurements is received. Therefore, there is no need to wait till sufficient number of measurements are collected to start the algorithm. The simulation results of illustrative examples are presented to show the effectiveness of the proposed estimation scheme.

Keywords: minimum variance estimator; constrained parameters; recursive estimation; system identification; linear systems; dynamic systems; discrete time systems; parameter estimation; simulation.

DOI: 10.1504/IJCAT.2011.045817

International Journal of Computer Applications in Technology, 2011 Vol.42 No.4, pp.267 - 278

Published online: 12 Mar 2012 *

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