Skip to Main content Skip to Navigation
New interface

Identification des systèmes en boucle fermée. Contributions aux méthodes de compensation de biais et des sous-espaces

Abstract : The thesis deals with closed-loop identification of SISO or MIMO systems, represented by a linear time-invariant-model either in transfer or state-space forms. This work is divided into three main parts. The first part gives key issues associated with closed-loop system identification as well as a bibliographic synthesis of the different available methods, classicall divided into three broad categories and referred to as direct, indirect and joint inout/output approaches. The second part focuses on the so-called bias-eliminated least-squares method for closed-loop identification of a transfer function by an indirect approach. It basically consists in estimating and then removing the bias introduced by the least-squares algorithm. The contribution of this chapter is twofold. At first proposition aims as enlightening the interpretation of those bias eliminated techniques by demonstrating their membership to the instrumental variable estimator class. A second proposition develops an extension of the previous method to the closed-loop continuous-time system identification, from sampled input/output data. Then, the last part of this thesis adresses the crucial (open) point of analyzing and assessing the quality of a nominal closed-loop state-space model stemmed from a subspace technique. In this respect, the identification algorithm is formulated in terms of a criterion minimisaton. This formulation is used on the one hand, to develop an algorithm improvement for estimating the nominal model. On the second hand, several methofd for estimating uncertainty regions of nominal model invariant parameters are proposed.
Document type :
Complete list of metadata
Contributor : Thèses UL Connect in order to contact the contributor
Submitted on : Thursday, March 29, 2018 - 10:38:32 AM
Last modification on : Saturday, October 16, 2021 - 11:14:14 AM

Intranet access


  • HAL Id : tel-01746473, version 1



Marion Gilson. Identification des systèmes en boucle fermée. Contributions aux méthodes de compensation de biais et des sous-espaces. Autre. Université Henri Poincaré - Nancy 1, 2000. Français. ⟨NNT : 2000NAN10119⟩. ⟨tel-01746473⟩



Record views