摘要
根据文献[1],测井资料最优化解释方法的目标是求取一个在最小二乘意义下拟合测井数据的解释模型。但是,由此得到的解释结果能在多大程度上反映地层参数的实际情况,即对求解结果如何进行定量评价,这是理论上需要解决的重要问题之一。本文试图予以初步探讨。 设(?)是求解结果,CR_α是以(?)为中心的置信区间,(?)是地层参数的真值,定义为真值(?)属于置信区间CR_α的概率。本文详细论讨了CR_α的求法,并用蒙特卡罗模拟作了检验。本文的结论也适合于地球物理中其它基于最小二乘的反演问题。
Based on reference [1], the purpose of optimal well logging interpretation is to build up a model used for interpretation. The practical procedure satisfies the least square principle and the interpretation is by fitting the well log data to this model. But how well does the final interpretation reflect the formation parameters, or in other words, how to evaluate the result obtained quantitatively, this is indeed an important problem needed to be solved in theory. Also, this is the key problem of this
paper.
Supposing x is the result extracted, CRα is the confidence interval
with x as its centre, x is the true value of the formation, then, we
can define P[x∈CRα]=α to be the probability of x within the confidence interval CR@.
In this paper, the extraction of CRα is examined and it has been verified by the Monte Carlo simulation. Conclusions reasoned out are suitable for geophysical inversion based on the least square principle.
出处
《石油物探》
EI
CSCD
北大核心
1989年第2期64-72,共9页
Geophysical Prospecting For Petroleum
