.png){kind=icon}
.png){kind=icon}
Return
A general framework of Bayesian network for system reliability analysis using junction tree
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ress.2021.107952.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
To perform the reliability analysis of complex and large-scale systems, Bayesian network (BN) can be useful as it facilitates modelling the causal relationship between multiple types of variables, e.g. hazards, material properties, and inspection results. However, its conventional approach shows limitations in handling large-scale systems and advanced inference tasks such as continuous distributions and approximate inference. On the other hand, these issues have been successfully addressed by system reliability analysis (SRA) theory, while the complexity of system reliability methods (SRMs) makes it challenging to handle multiple types of variables collectively. Accordingly, to facilitate the reliability analysis of real-world problems, this paper develops a general framework to implement BN for SRA by employing junction tree (JT). The connection between BN and SRA is further consolidated by summarizing common computational challenges and proposing heuristics to resolve them. While it provides a systematic way to implement SRMs within the BN framework, such generalization can also be used to enhance the functionality of the general-purpose software programs developed for BN as demonstrated by the companion Matlab (R)-based toolkit BNS-JT. The applicability and efficiency of the proposed framework are demonstrated by numerical examples.
Keywords:
Bayesian network
system reliability analysis
junction tree
complex system
large-scale system
hybrid distribution
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
R
IF:
11
Papers:
9.0K
Citations:
4.2W