System reliability analysis method based fuzzy probability

Li Zhi-Gang,Zhou Jun-Gang,Liu Bo-Ying,Li Zhi-gang,Zhou Jun-gang,Liu Bo-ying
DOI: https://doi.org/10.1109/ifuzzy.2016.8004956
2016-11-01
Abstract:Traditional methods for analyzing system reliability involve the use of unit reliability and system composition. Probability theory-based calculations are subsequently made to obtain system reliability. However, these methods fail to consider the type of failure distribution that exists in system units, resulting in inaccurate system reliability calculations at any given time. For two systems that feature the same system reliability, these methods also fail to take into account the effect of differences in unit reliability on system reliability. The process in which system units go from a normal working state to a failure state can be described as a fuzzy process. In this study, introduced a novel method that based on fuzzy probability analyze system reliability. This method is based on the assumption that unit reliability is a fuzzy probability-based event and involves analyzing failure distribution membership function to determine the reliability of system units at any given time. The concepts of system failure probability (which is a characteristic quantity used to measure the effect of differences in unit reliability on system reliability) and integrated system reliability (which is a characteristic quantity that allows comprehensive evaluation of system reliability) are introduced. The proposed method was adopted in the reliability analysis of series, parallel, and compound systems. The results verified the effectiveness of the proposed method.
What problem does this paper attempt to address?