Gas turbine fault diagnosis using fuzzy-based decision fusion

see the original item page
in the repository's web site and access all digital files if the item*



Gas turbine fault diagnosis using fuzzy-based decision fusion (EN)

Mathioudakis, K (EN)
Kyriazis, A (EN)

journalArticle (EN)

2014-03-01T01:30:47Z
2009 (EN)


A two-step information fusion technique allowing the combination of fast response and performance data for the improvement of gas turbines diagnostic procedures is proposed. The proposed technique is derived from the notion of decision level fusion. Different diagnostic methods provide assessments for the condition of an engine, and the final decision is derived from a combination of these assessments. The diagnostic method of probabilistic neural networks acts independently and in parallel on data of a different nature. The conclusions are given for the first step of the fusion technique and are aggregated deriving the probability consensus. In a second step this consensus is classified within a set of examined faults using the fuzzy set theory and fuzzy logic, thus providing the final diagnostic decision. The effectiveness of the proposed technique is demonstrated through the application to data from a radial and an axial compressor. Copyright © 2008 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved. (EN)

Engineering, Aerospace (EN)

Gas Turbine (EN)
Fault Diagnosis (EN)
Probabilistic neural networks (EN)
Diagnostic decisions (EN)
Final decision (EN)
Fusion techniques (EN)
Turbomachinery (EN)
Diagnostic methods (EN)
Decision level fusion (EN)
Fast response (EN)
Fault diagnosis (EN)
Fuzzy logic (EN)
Fuzzy sets (EN)
Decision Fusion (EN)
Gas turbines (EN)
Axial compressors (EN)
Decision theory (EN)
Neural networks (EN)
Information fusion techniques (EN)
Diagnostic procedure (EN)
Decision fusion (EN)
Indexing (of information) (EN)
Performance data (EN)

Journal of Propulsion and Power (EN)

English

AMER INST AERONAUT ASTRONAUT (EN)




*Institutions are responsible for keeping their URLs functional (digital file, item page in repository site)