Skip to Main content Skip to Navigation
Habilitation à diriger des recherches

Estimation, diagnostic et commande tolérante de systèmes décrits par des multimodèles

Abstract : The works described in this thesis deal with state estimation, fault diagnosis and fault tolerant control of complex systems represented by nonlinear models. The complexity may be due to the nature of the model equations (differential, nonlinear, etc), to the uncertainties affecting these relations (unknown inputs or parameters) or to the melting of different classes of equations. The studied models are the Takagi-Sugeno (TS) models and the switched or T-S singular models. These formalisms allow to accurately describe non-linearities, multiple time scales, simultaneous discrete and continuous phenomena, etc. Regarding the T-S systems, consisting in the interpolation between a set of linear systems, a special attention is paid to the case of state dependent interpolation functions. Different design algorithms of state observers are exposed and then used for diagnosis by direct fault measurement or structured residual generation. The fault and state estimates are then exploited for the synthesis of fault and input-saturation tolerant control laws. State and unknown input estimation of nonlinear singular systems is also addressed.
Document type :
Habilitation à diriger des recherches
Complete list of metadata

Cited literature [205 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/tel-01418847
Contributor : Benoit Marx Connect in order to contact the contributor
Submitted on : Saturday, December 17, 2016 - 1:31:15 PM
Last modification on : Thursday, November 25, 2021 - 10:56:09 AM
Long-term archiving on: : Monday, March 20, 2017 - 7:52:25 PM

Identifiers

  • HAL Id : tel-01418847, version 1

Collections

Citation

Benoît Marx. Estimation, diagnostic et commande tolérante de systèmes décrits par des multimodèles. Automatique / Robotique. Université de Lorraine, 2016. ⟨tel-01418847⟩

Share

Metrics

Les métriques sont temporairement indisponibles