Skip to Main content Skip to Navigation
Theses

Utilisation des techniques de redondances matérielles et analytiques à la détection de pannes de capteurs. Application aux centrales nucléaires

Abstract : The study of failure detection concentrates on the folLowing points : - detection strictly speaking, - location the defective sensor, - correction of the error. The efficiency of a few methods applied to steady-state analytic redundancy, whether linearor not, is evaluated (parity space, standardized imbalance residuals approach, standardized least-square residuals analysis) using simulated data. The use of non-linear models for analytical redundancy is examined. From the standardized imbalance residuals approach (SIRA) and the standardized least-square residuals analysis (SLSRA), we present an extension to non-linear models. vVe dwell on the mathematical properties of the non-linear models. We show the problem of the computation of the estimate of the true values vector and the difficulties that we meet with fusion of not necessarily linear equations. The SLSRA is essentially examined with constraint linearization.The SLSRA method with constraint linearization was applied to the nstrumentation of a stearri generator in the Paluel PWR plant. Results are promising.
Document type :
Theses
Complete list of metadata

https://hal.univ-lorraine.fr/tel-01747485
Contributor : Thèses UL Connect in order to contact the contributor
Submitted on : Thursday, March 29, 2018 - 11:01:15 AM
Last modification on : Monday, December 13, 2021 - 10:49:31 AM
Long-term archiving on: : Friday, September 14, 2018 - 12:38:30 AM

File

SCD_T_1991_0072_KRATZ.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-01747485, version 1

Collections

Citation

Frédéric Kratz. Utilisation des techniques de redondances matérielles et analytiques à la détection de pannes de capteurs. Application aux centrales nucléaires. Sciences de l'ingénieur [physics]. Université Henri Poincaré - Nancy 1, 1991. Français. ⟨NNT : 1991NAN10072⟩. ⟨tel-01747485⟩

Share

Metrics

Record views

78

Files downloads

1127