The thesis project, developed in collaboration with Leonardo S.p.A. Helicopters Division Company, presents a procedure to analyze events that occurred on an aircraft fleet following a failure of a system or of a component. The procedure evaluates the hazard rate and other statistical reliability functions of the component or system, then, it identifies the presence of an unsafe or not unsafe condition. If an unsafe condition is detected, an adequate reaction time and a corrective action will be identified in accordance with relevant EASA regulation rules. The aim of this project has been the development of an algorithm able to perform the above mentioned analysis. The first section of this thesis is a synthesis of the certification framework with which any European aeronautical company operates. The regulation of interest is the EASA Part-21 Annex I. The second section focuses on the analysis of mechanical failures in the aviation industry. This analysis is carried out through reliability control procedures used in manufacturing. Component times to failure are assumed to have a Weibull distribution since the Weibull function has demonstrated the ability to give a good fitting of real life data, especially in the aviation industry. Its main properties will be presented. The distribution parameters are estimated with four different methods: the Median Rank Regression (MRR), the Maximum Likelihood Estimation (MLE), MLE with Reduced Bias Adjustment (RBA) and Weibayes method. The third section describes the algorithm implemented to compute the failure rate analysis and the rectification time. The performances of the different methods have been compared to identify the method able to provide the best fitting of the real data. In the last section, the results of two different case studies are reported.
Il lavoro di tesi, svolto in collaborazione con l’azienda Leonardo S.p.A. Helicopters division, presenta una procedura per l’analisi di avarie meccaniche all’interno di una flotta aeronautica e del loro tasso di guasto. In merito ad un’avaria, l’evento relativo viene classificato come unsafe o not unsafe in base a regole imposte dall’autorità Europea EASA. Se l’evento è identificato come unsafe, viene effettuato il calcolo dell’intervallo di tempo concesso dall’ente regolativo entro cui adottare le azioni correttive sulla flotta a rischio. Scopo del progetto è stata l’implementazione di un algoritmo in grado di eseguire l’analisi richiesta. Nella prima sezione, vengono sinteticamente riportate le linee guida dell’ente certificativo EASA, documento Part-21 Annex I. I capitoli successivi sono incentrati sull’analisi di avarie meccaniche in campo aeronautico. Si suppone che i tempi associati alle rotture dei componenti seguano una distribuzione di Weibull, in quanto la funzione di Weibull ha dimostrato di approssimare con buona accuratezza i dati reali, soprattutto nel campo dell’aviazione. Vengono quindi descritte le principali caratteristiche di tale distribuzione statistica. La stima dei parametri è stata effettuata attraverso metodologie differenti: Median Rank Regression (MRR), Maximum Likelihood Estimation (MLE), Maximum Likelihood Estimation (MLE) con Reduced Bias Adjustment (RBA) e metodo di Weibayes. Dopo aver definito l’approccio e i metodi utilizzati, è stato implementato l’algoritmo in grado di eseguire, dati avarie e flotta, l’analisi dei tassi di guasto e il calcolo del tempo di rettificazione in caso di condizione unsafe. Un confronto fra i risultati di MRR, MLE, MLE-RBA e metodo di Weibayes ha permesso di scegliere il procedimento in grado di approssimare al meglio i dati reali. Nell'ultima parte della tesi, sono stati riportati i risultati di due casi studio.
Development of an algorithm for statistical analysis of aircraft fleet events
D'ASCENZO, ANNA
2017/2018
Abstract
The thesis project, developed in collaboration with Leonardo S.p.A. Helicopters Division Company, presents a procedure to analyze events that occurred on an aircraft fleet following a failure of a system or of a component. The procedure evaluates the hazard rate and other statistical reliability functions of the component or system, then, it identifies the presence of an unsafe or not unsafe condition. If an unsafe condition is detected, an adequate reaction time and a corrective action will be identified in accordance with relevant EASA regulation rules. The aim of this project has been the development of an algorithm able to perform the above mentioned analysis. The first section of this thesis is a synthesis of the certification framework with which any European aeronautical company operates. The regulation of interest is the EASA Part-21 Annex I. The second section focuses on the analysis of mechanical failures in the aviation industry. This analysis is carried out through reliability control procedures used in manufacturing. Component times to failure are assumed to have a Weibull distribution since the Weibull function has demonstrated the ability to give a good fitting of real life data, especially in the aviation industry. Its main properties will be presented. The distribution parameters are estimated with four different methods: the Median Rank Regression (MRR), the Maximum Likelihood Estimation (MLE), MLE with Reduced Bias Adjustment (RBA) and Weibayes method. The third section describes the algorithm implemented to compute the failure rate analysis and the rectification time. The performances of the different methods have been compared to identify the method able to provide the best fitting of the real data. In the last section, the results of two different case studies are reported.| File | Dimensione | Formato | |
|---|---|---|---|
|
2018_04_DAscenzo.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
1.5 MB
Formato
Adobe PDF
|
1.5 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/139500