摘要
本文讨论了测井资料最优化解释方法质量控制的含义和具体应用,指出:(1)置信度曲线[a_i-√σ_i^2 + τ_i^2,a_i + √σ_i^2 + τ_i^2]用于检验假设a_i - f_i(X,S)～N(0,σ_i^2 + τ_i^2),i=1,2,…,n;(2)减小非相关函数值Rinc则用于检验假设△(a,X)～X^2(n + t - m)。计算表明:P[|a_i - f_i(X,S)|≤(σ_i^2 + τ_i^2)^(1/2)]68.3%,i=1,2,…,n,以及P [R_(inc)~2≤1]=84.2%。
This paper discusses the defination and application of quiality control in optimum log interpretation technique, and shows: (1) the confidence
curve [a_i-√/σ_i^2 + τ_i^2, a_i + √σ_i^2 + τ_i^2] is used to confirm the assumption
a_i-f_i (X, s)～N (0, σ_i^2 + τ_i^2), i = 1, 2,…, n; (2) reduced incoherence function Rinc
is to confirm assumption △(a, X)～x^2 (n + t - m). The computation shows:
P[|a_i-f_i(X, S)|≤(σ_i^2+τ_i^2)^(1/2)= 68.3%, i=1, 2,…n, and P [Rinc^2≤1]= 84.2%.
