In systems with Imperfect Fault Coverage (IFC), all components are subject to uncovered failures, possibly threatening the whole system. Therefore, to improve the system reliability, it is important to timely detect, identify, and shut down the components that are no more relevant for the system operation. Thus, the Irrelevance Coverage Model (ICM) was proposed based on the Imperfect Fault Coverage Model (IFCM). In the ICM, any component detected as irrelevant can be safely shut down without reducing the system reliability and preventing the case where its eventual failure may remain uncovered and cause a direct system failure. This not only improves the system reliability but also saves energy. This thesis solves the problem of quantitative evaluation of component relevance. It assumes that components have independent and identically distributed~(i.i.d.) lifetimes to describe only the impact of the system design on the system reliability and energy consumption. To this end, the Component Relevance is proposed to represent the probability that a component can keep its relevance throughout the system lifetime. Then, the Birnbaum Importance (BI) measure is applied to the system with ICM. The BI measure with ICM considers the relevance of the components while considering the reliability of the components. At the same time, the changes of the importance of the components in three different models, i.e., Perfect Fault Coverage Model (PFCM), Imperfect Fault Coverage Model (IFCM), and ICM, are analyzed. Moreover, the Dynamic Relevance Measure~(DRM) is defined to characterize the irrelevant components in different stages of the system lifetime depending on the number of occurred component failures, supporting the evaluation of the probability that the system fails due to uncovered failures of irrelevant components. Also, the gain in shutting down the irrelevant components in the ICM can be evaluated both in terms of the energy saved and the fraction of the average system lifetime during the system is not coherent. Finally, the system reliability over time is also efficiently derived, both in the case that irrelevance is not considered and in the case that irrelevant components can be immediately isolated, notably supporting any general (i.e.,~non-Markovian) distribution for the failure time of components. The feasibility and effectiveness of the proposed analysis methods are assessed on two real-scale case studies addressing the reliability evaluation of a flight control system and a multi-hop Wireless Sensor Network~(WSN). I have obtained the most important components for the left edge flap of the F18 flight control system to improve the component reliability, which improves system reliability more obviously. For the different topologies of WSN, the reliability and relevance of the Diagonal topology are better than the Orthogonal topology. So the WSN with Diagonal topology should be given priority in the system phase.

Analysis of the Components and Systems Relevance / Luyao Ye. - (2021).

Analysis of the Components and Systems Relevance

Luyao Ye
2021

Abstract

In systems with Imperfect Fault Coverage (IFC), all components are subject to uncovered failures, possibly threatening the whole system. Therefore, to improve the system reliability, it is important to timely detect, identify, and shut down the components that are no more relevant for the system operation. Thus, the Irrelevance Coverage Model (ICM) was proposed based on the Imperfect Fault Coverage Model (IFCM). In the ICM, any component detected as irrelevant can be safely shut down without reducing the system reliability and preventing the case where its eventual failure may remain uncovered and cause a direct system failure. This not only improves the system reliability but also saves energy. This thesis solves the problem of quantitative evaluation of component relevance. It assumes that components have independent and identically distributed~(i.i.d.) lifetimes to describe only the impact of the system design on the system reliability and energy consumption. To this end, the Component Relevance is proposed to represent the probability that a component can keep its relevance throughout the system lifetime. Then, the Birnbaum Importance (BI) measure is applied to the system with ICM. The BI measure with ICM considers the relevance of the components while considering the reliability of the components. At the same time, the changes of the importance of the components in three different models, i.e., Perfect Fault Coverage Model (PFCM), Imperfect Fault Coverage Model (IFCM), and ICM, are analyzed. Moreover, the Dynamic Relevance Measure~(DRM) is defined to characterize the irrelevant components in different stages of the system lifetime depending on the number of occurred component failures, supporting the evaluation of the probability that the system fails due to uncovered failures of irrelevant components. Also, the gain in shutting down the irrelevant components in the ICM can be evaluated both in terms of the energy saved and the fraction of the average system lifetime during the system is not coherent. Finally, the system reliability over time is also efficiently derived, both in the case that irrelevance is not considered and in the case that irrelevant components can be immediately isolated, notably supporting any general (i.e.,~non-Markovian) distribution for the failure time of components. The feasibility and effectiveness of the proposed analysis methods are assessed on two real-scale case studies addressing the reliability evaluation of a flight control system and a multi-hop Wireless Sensor Network~(WSN). I have obtained the most important components for the left edge flap of the F18 flight control system to improve the component reliability, which improves system reliability more obviously. For the different topologies of WSN, the reliability and relevance of the Diagonal topology are better than the Orthogonal topology. So the WSN with Diagonal topology should be given priority in the system phase.
2021
Enrico Vicario
REPUBBLICA POPOLARE CINESE
Luyao Ye
File in questo prodotto:
File Dimensione Formato  
Thesis-Luyao YE.pdf

Accesso chiuso

Tipologia: Tesi di dottorato
Licenza: Tutti i diritti riservati
Dimensione 2 MB
Formato Adobe PDF
2 MB Adobe PDF   Richiedi una copia

I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificatore per citare o creare un link a questa risorsa: https://hdl.handle.net/2158/1251754
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact