摘要
在缺乏足够数据或信息不完整的情况下 ,给出失效概率的界限或范围 ,比给出一确定的单值具有更高的可信度。文中提出概率模型的分布参数为模糊变量时结构模糊失效概率的计算方法。不同于一般的模糊失效概率模型 ,文中给出的失效概率本质上是模糊的。可根据模糊参数的可能性分布给出模糊失效概率的可能性分布。算例分析表明 ,当概率模型参数为模糊变量时 ,结构的失效概率具有较大的分散性。说明还模糊概率以其“模糊”的本来面目 ,可为分析和设计决策提供更全面、更真实的有用信息。
In reliability computation, when insufficient accurate data are available or available information is incomplete, a range representation of failure probability may be more confident than represented by a single number. Uncertainties of the parameters of probabilistic reliability model were described by fuzzy variables. A procedure was presented for computing the fuzzy probability of failure of mechanical structures, when parameters of probability distribution are fuzzy. By our method, possibility distribution of the fuzzy probability of failure may be given by the possibility distribution of the fuzzy parameters, which is different from some commonly used computing procedures of fuzzy probability. A numerical example show that, when probability parameters are fuzzy, the failure probability may has large disperse. So, the failure probability should be taken as a fuzzy variable to represent the safety of structures reasonably. It was also explained that more completely and really useful information may be offered for decision-making of analysis and design of structures when the essentially fuzziness of fuzzy probability are taken into account.
出处
《机械强度》
CAS
CSCD
北大核心
2003年第5期527-529,共3页
Journal of Mechanical Strength
基金
国家自然科学基金 (595750 4 0
597750 32 )
航空基金 (0 0B530 1 0 )资助项目~~
关键词
概率模型
概率参数
模糊变量
模糊失效概率
计算方法
Probabilistic model
Distribution parameter
Fuzzy variable
Fuzzy probability of failure
