Gas turbine diagnostic method evaluation using the ProDiMES software

Κοσκολέτος, Αναστάσιος - Ορέστης; Koskoletos, Anastasios - Orestis

dc.contributor.author	Κοσκολέτος, Αναστάσιος - Ορέστης	el
dc.contributor.author	Koskoletos, Anastasios - Orestis	en
dc.date.accessioned	2017-09-11T10:36:57Z
dc.date.issued	2017-09-11
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/45565
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.13871
dc.rights	Default License
dc.subject	Διαγνωστική	el
dc.subject	Αεριοστρόβιλοι	el
dc.subject	ProDiMES	en
dc.subject	Gas turbines	en
dc.subject	Diagnostics	el
dc.title	Gas turbine diagnostic method evaluation using the ProDiMES software	en
heal.type	bachelorThesis
heal.secondaryTitle	Αξιολόγηση διαγνωστικών μεθόδων αεριοστροβίλων με χρήση του λογισμικού ProDiMES	el
heal.classification	Διαγνωστική αεριοστροβίλων	el
heal.classification	Gas turbine diagnostics	en
heal.dateAvailable	2018-09-10T21:00:00Z
heal.language	el
heal.language	en
heal.access	campus
heal.recordProvider	ntua	el
heal.publicationDate	2017-07-07
heal.abstract	In the present thesis, the development and performance evaluation of six gas turbine diagnostic methods using NASA’s Propulsion Diagnostic Method Evaluation Strategy (ProDiMES) software is presented. ProDiMES is a public benchmarking diagnostic problem that also provides a set of evaluation metrics to enable the comparison of candidate aircraft engine gas path diagnostic methods. In the course of its useful life an engine is susceptible to encountering a wide variety of physical problems. These include such things as erosion, corrosion, fouling, excessive tip clearances, worn seals, plugged fuel nozzles, faulty sensors, etc. The main purpose of a gas path diagnostic method is to detect as many as possible of these problems through the observation of available or appropriately selected sensed parameters (measurements). The detectability of these problems depends on their nature, magnitude and changes they cause in the engine’s sensed parameters. Since these problems can cause significant performance shifts in an engine, their early and accurate identification is of crucial importance. In the past decades, many diagnostic techniques have been proposed for diagnosing engine component and sensor faults, including linear and nonlinear gas path analysis methods, artificial neural network, fuzzy logic, genetic algorithms and expert systems as well as hybrid methods. The diagnostic methods to be evaluated in the present thesis are the following: Probabilistic Neural Network (PNN), k-Nearest Neighbor (k-NN), Combinatorial, Optimization, Adaptive 2x2, Search. All of the above methods, except for the k-Nearest Neighbor, have been proposed by the Laboratory of Thermal Turbomachinery of the National Technical University of Athens (NTUA/LTT). Part of this thesis was to programmatically develop these methods in the Matlab environment. In addition, a Windows application, namely C-MAPSS Steady State Diagnostic Tool (CDT), was developed in Microsoft’s Visual Studio and coded in Visual Basic, where all of the aforementioned diagnostic methods are implemented.	en
heal.advisorName	Αρετάκης, Νικόλαος	el
heal.committeeMemberName	Μαθιουδάκης, Κωνσταντίνος	el
heal.committeeMemberName	Γιαννάκογλου, Κυριάκος	el
heal.committeeMemberName	Αρετάκης, Νικόλαος	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Μηχανολόγων Μηχανικών. Τομέας Ρευστών. Εργαστήριο Θερμικών Στροβιλομηχανών	el
heal.academicPublisherID	ntua
heal.numberOfPages	148 σ.
heal.fullTextAvailability	true