This paper presents a method which aims at improving parameter estimation i
n dynamical systems. The general principle of the method is based on a modi
fication of the least-squares objective function by means of a weighting op
erator, in view to improve the conditioning of the identification problem.
First we recall a previous work using variational calculus in order to obta
in the weighting operators through a linear equation. Then we propose a new
approach which consists of determining the weights by formulating an optim
ization problem including positive semidefinite constraints (linear matrix
inequalities, LMI). Copyright (C) 2001 John Wiley & Sons, Ltd.