| In this paper, a commonly used method to analyze fuzzy reliability by transforming the problem of fuzzy reliability to the problem of conventional reliability via the cut-set conception of fuzzy sets was introduced firstly. Secondly, based on the method mentioned above, the methods of establishing fuzzy event membership function and of how to use the established membership function to solve the problem of fuzzy reliability were studied when one of stress and strength is a fuzzy variable and the other is a random variable. Thirdly, the paper discussed a method to estimate fuzzy reliability based on the experience data, and obtained confidence intervals by using the above method. Fourthly, the form of the fuzzy safety/failure event membership function was obtained when the stress or strength was a normal fuzzy variable. Therefore, in order to establish a simple and unified analytical method of fuzzy reliability, non-normal fuzzy variables are transformed to normal fuzzy ones and the equivalent normal membership functions transformed from some symmetrical non-normal membership functions are derived. Lastly, the parametric sensitivity in fuzzy reliability analysis was extensively discussed and a computational method of the parametric sensitivity was presented when the membership function of fuzzy variable is a linear or normal one.The research of this paper puts forward a uniform method based on fuzzy event probability theory, and provides a new thought for fuzzy reliability analysis. Also, the method given in the paper is a better one than anyone before.
|
PDF Full Text Request