Skip to Main content Skip to Navigation

Validation de données et diagnostic des systèmes incertains à l'aide de l'analyse par intervalles

Abstract : To function correctly, the control and the monitoring systems of the processes require to receive consistent data representative of the operating condition of these processes. Unfortunately, the data taken on a process does not constitute an exact representation of its operation because the data are subject to errors of various natures. Therefor, it is necessary to test the validity of the data gathered before using them. The major problem encountered at the time of a procedure of data validation or diagnosis is that a model of a process defines only one approximate behavior of the process which it is supposed to represent. These approximations come from the errors of modeling which can be explained by the impossibility of identifying exactly the parameters of a real system because of the limited precision of its instrumentation. Moreover, the behavior of any real system changes in the time in a nonforeseeable way because of the environment, of the procedures... The representation of a real system by an uncertain model constitutes a good solution to take into account the uncertain character of the parameters model of the system. Several approaches were developed to analyse such models, as the probabilistic approach and the bounding approach. The objective of this thesis is to propose, by using the bounding approach, a methods of data validation for the uncertains systems.
Document type :
File URL :
Complete list of metadata
Contributor : Thèses Ul <>
Submitted on : Thursday, March 29, 2018 - 12:34:49 PM
Last modification on : Thursday, March 12, 2020 - 3:08:35 PM


  • HAL Id : tel-01749969, version 1



Moustapha Alhaj-Dibo. Validation de données et diagnostic des systèmes incertains à l'aide de l'analyse par intervalles. Autre [cs.OH]. Institut National Polytechnique de Lorraine, 2005. Français. ⟨NNT : 2005INPL058N⟩. ⟨tel-01749969⟩



Record views