Approximation of Ideal Fault Detector in Real-time FDI with Supervised Training
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21220%2F06%3A02120292" target="_blank" >RIV/68407700:21220/06:02120292 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—
Alternative languages
Result language
angličtina
Original language name
Approximation of Ideal Fault Detector in Real-time FDI with Supervised Training
Original language description
The paper generally attempts to propose the algorithm, which helps to approximate in real time the most likely system mode in case of "new" data information. An approximation is performed in neighbourhoods around the actual observed data vector. Besides,the suggested algorithm, which is a part of the Bayesian identification-based real-time fault detection and isolation process, "fights" with the dimensionality problem connected with use of probabilistic models in FDI. The practical aspects of the proposed algorithm have been tested in real time on a laboratory heating system. Results obtained proved suitability of the designed algorithm for the approximation of ideal fault detector.
Czech name
Aproximace ideálního detektoru chyb při diagnostice chyb v reálném čase s dohledem trénování
Czech description
Článek navrhuje algoritmus, který pomocí pravděpodobnostního diskrétního modelu provádí detekci a lokalizaci chyb nelineárního stochastického dynamického systému. Navržená metoda pracující v reálném čase umožňuje průběžně doplňovat nové informace o chybách.
Classification
Type
D - Article in proceedings
CEP classification
BC - Theory and management systems
OECD FORD branch
—
Result continuities
Project
—
Continuities
S - Specificky vyzkum na vysokych skolach
Others
Publication year
2006
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Advanced Engineering Design AED 2006
ISBN
80-86059-44-8
ISSN
—
e-ISSN
—
Number of pages
8
Pages from-to
—
Publisher name
ČVUT
Place of publication
Praha
Event location
Prague
Event date
Jun 11, 2006
Type of event by nationality
EUR - Evropská akce
UT code for WoS article
—