关键词: 结构可靠度, 置信下限, 样本空间排序法

Abstract: Let X₁, X₂,…，X_n be independent and identically distributed random variables. X₁ is distributed as N(μ, σ²), where μ and σ are unknown parameters satisfying μ>0 and σ/μ≤c₀(c₀ is known). Given t_i>0(i=1,…,n). Suppose that the observation data is Z=(Z₁,…，Z_n)，in which Z_i=1 or 0 according as X_i is bigger than t_i or not. R(t₀)=P(X₁>t₀) is so-called structure reliability. A method seeking the lower confidence limit R_L of R(t₀) based on the data Z is given. Especially, the accurate computational method is obtained in the case of no failure (i.e.Z_i≡1).

Key words: structure reliability, lower confidence limit, ordering method of the sample space