Parameter estimation algorithms and robust identifiability

Abstract

Parameter estimation algorithms are obtained for a class of finite-dimensional systems. The algorithms are presented in a completely deterministic setting and their convergence properties are investigated by means of a fixed-point theorem. The concept of robust identifiability is introduced. The invertibility of a certain matrix is shown to be a necessary condition for robust identifiability. Some numerical results are presented.

Additional information

Notes on contributors

G. A. MEDRANO-CERDA‡

‡Now at the Department of Electronic and Electrical Engineering, University of Salford, Salford M5 4WT, U. K.