Contribution à l'estimation de l'état des systèmes singuliers : application à la validation de données des systèmes dynamiques linéaires

Abstract : The purpose of this thesis is the study of discrete linear singular systems and data reconciliation of dynamic linear systems. Data reconciliation of steady state linear systems is tackled in the first part. The state of these systems is estimated in the maximum likelihood sense, the variance of the measurement errors being either known or unknown and diagonal. The second part is devoted to the study of discrete linear singular systems which have a number of unknowns superior or equal to the number of equations. These systems have either a process noise or not. The state of these systems is recursively estimated with the maximum likelihood estimator. The convergence of these state estimation algorithms is studied. If there is not a process noise, the state estimation algorithm is factorized. If there is a process noise, the state estimation algorithm corresponds to a generalization of the Kalman filter and is adapted to the time-varying case. In the last part, the state estimation algorithms are used for systems described by linear dynamic balance equations and for simultaneous estimation of state variables and inputs of standard state-space systems.
Document type :
Theses
File URL :
http://docnum.univ-lorraine.fr/prive/SCD_T_1990_0405_ZASADZINSKI.pdf
Complete list of metadatas

https://hal.univ-lorraine.fr/tel-01754300
Contributor : Thèses Ul <>
Submitted on : Friday, March 30, 2018 - 9:44:05 AM
Last modification on : Saturday, March 31, 2018 - 1:24:39 AM

Identifiers

  • HAL Id : tel-01754300, version 1

Collections

Citation

Michel Zasadzinski. Contribution à l'estimation de l'état des systèmes singuliers : application à la validation de données des systèmes dynamiques linéaires. Autre [cs.OH]. Université Henri Poincaré - Nancy 1, 1990. Français. ⟨NNT : 1990NAN10405⟩. ⟨tel-01754300⟩

Share

Metrics

Record views

11