Sequential correlation-based propagator algorithm for recursive subspace identification
| Authors: | Mercère Guillaume, Laboratoire d'Automatique, Génie Informatique et Signal, France
Lecoeuche Stéphane, Laboratoire d'Automatique, Génie Informatique et Signal, France
Vasseur Christian, Laboratoire d'Automatique, Génie Informatique et Signal, France |
|---|
| Topic: | 1.1 Modelling, Identification & Signal Processing |
|---|
| Session: | Subspace Methods |
|---|
| Keywords: | Identification algorithms, recursive algorithms, subspace methods, state-space models, recursive least squares |
|---|
Abstract
The problem of MIMO state space recursive identification in the framework of subspace methods is considered in this article. Two new recursive algorithms, based a recursive formulation of the \texttt{MOESP} identification class, are more precisely developed. The specific feature of these methods is that they share an only algorithm to recursively estimate a basis of the observability matrix. Two propagator based \cite{MD91} criteria are more precisely introduced. A sequential RLS algorithm is then proposed to equally minimise these cost functions. The benefits of these new methods in comparison with \texttt{EIVPAST} and \texttt{UDPAST} \cite{LGV00} are stressed on a simulation example. \textit{Copyright}$^{\mathit{\copyright}}$\textit{2005 IFAC}.
