The objective of this research is to propose a Fault Detection and Isolation (FDI) system based on data-driven approach. The choice of approaching the fault diagnosis problem with a model-free methodology rather than a model-based approach was motivated by the fact that in some particular applications dynamic models are not available or not appropriate for fault detection and isolation purposes. The proposed innovative FDI system combines Principal Component Analysis (PCA), Cluster Analysis and Pattern Recognition technique: the result is the so called Fuzzy Faults Classifier (FFC). The combination of these techniques allows to automatically detect and isolate single and multiple faults and allow to overcome the growth of the complexity in the analysis of process faults that typically involve many variables. The research also proposes the use of the adaptive thresholds: the thresholds scheme follows a classical structure proposed in literature but the parameters used on it have been computed by an innovative approach based on the spectral analysis of the process inputs. The main task of PCA, i.e. the choice of the Principal Component retained in the model, has been accomplished by a new method, based on the statistical test ANOVA (ANalysis Of VAriance) and a comparison with other criteria present in literature has been made. The FDI system has been tested for the detection and the isolation of single and multiple faults as well as process faults of two particular systems: Multishaft Centrifugal Compressor and Unmanned Surface Vehicle (USV). The goodness and the efficiency of the proposed Fault Detection and Isolation system can be appreciate by the inspection of the results obtained in the real process. The results confirm the ability of the system in terms of fault detection and fault isolation and the possibility to extend its use to different real process. The only requirement is the presence of good measurement concerning the main process variables.
Lo scopo di questa tesi è la presentazione di un innovativo sistema di Fault Detection e Isolation (FDI) basato su un approccio data-driven. La scelta di utilizzare un approccio data-based è stata dettata dal fatto che, molto spesso, in applicazioni industriali, modelli dinamici del processo non sono disponibili o non sono particolarmente utili per scopi diagnostici. L’innovativo sistema FDI proposto combina le seguenti metodologie: Principal Component Analysis (PCA), Cluster Analysis e Pattern Recognition. Il risultato è quello che l’autore ha definito Fuzzy Fault Classifier (FFC). La combinazione delle tre tecniche sopracitate permette di rilevare ed isolare automaticamente guasti singoli e guasti multipli. Questo sistema è stato altresì arricchito dall’uso di soglie adattative la cui progettazione si è inspirata a degli schemi classici proposti in letteratura; l’apporto innovativo è stato quello di identificare alcuni parametri incogniti attraverso un approccio basato sull’uso dell’analisi spettrale dei segnali di ingresso. Per risolvere il cruciale problema della scelta delle Componenti Principali nella PCA, è stato presentato un nuovo metodo basato sul noto test statistico ANOVA (ANalysis Of VAriance); tale metodologia è stata confrontata con altri approcci ed i suoi benefici possono essere valutati. Il sistema FDI descritto è stato testato per valutare le capacità di identificazione e di isolamento dei guasti singoli e multipli, assieme ai principali gusti di processo, dei seguenti processi reali: Compressore Centrifugo Multistadio e Unmanned Surface Vehicle (USV). La bontà e l'efficienza del sistema FDI possono essere apprezzate valutando i risultati ottenuti. Essi confermano la capacità del sistema in termini di individuazione ed isolamento dei guasti ed inoltre, vista la sua caratteristica data-based, la possibilità di estenderne l'uso ad altri processi. L'unico requisito è la presenza di misure affidabili delle principali grandezze di processo.
Data-based design of fault detection and isolation (FDI) systems / Astolfi, Giacomo. - (2013 Jan 25).
Data-based design of fault detection and isolation (FDI) systems
Astolfi, Giacomo
2013-01-25
Abstract
The objective of this research is to propose a Fault Detection and Isolation (FDI) system based on data-driven approach. The choice of approaching the fault diagnosis problem with a model-free methodology rather than a model-based approach was motivated by the fact that in some particular applications dynamic models are not available or not appropriate for fault detection and isolation purposes. The proposed innovative FDI system combines Principal Component Analysis (PCA), Cluster Analysis and Pattern Recognition technique: the result is the so called Fuzzy Faults Classifier (FFC). The combination of these techniques allows to automatically detect and isolate single and multiple faults and allow to overcome the growth of the complexity in the analysis of process faults that typically involve many variables. The research also proposes the use of the adaptive thresholds: the thresholds scheme follows a classical structure proposed in literature but the parameters used on it have been computed by an innovative approach based on the spectral analysis of the process inputs. The main task of PCA, i.e. the choice of the Principal Component retained in the model, has been accomplished by a new method, based on the statistical test ANOVA (ANalysis Of VAriance) and a comparison with other criteria present in literature has been made. The FDI system has been tested for the detection and the isolation of single and multiple faults as well as process faults of two particular systems: Multishaft Centrifugal Compressor and Unmanned Surface Vehicle (USV). The goodness and the efficiency of the proposed Fault Detection and Isolation system can be appreciate by the inspection of the results obtained in the real process. The results confirm the ability of the system in terms of fault detection and fault isolation and the possibility to extend its use to different real process. The only requirement is the presence of good measurement concerning the main process variables.File | Dimensione | Formato | |
---|---|---|---|
Tesi.Astolfi.pdf
Solo gestori archivio
Tipologia:
Tesi di dottorato
Licenza d'uso:
Non specificato
Dimensione
5.11 MB
Formato
Adobe PDF
|
5.11 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.