【摘要】 《测量不确定度表示指南》(GUM)这个技术规范已被广泛认可,且其测量值关联一个不确定度值的建议也被广泛采纳。然而,这个规范遵循一种固有的概率方法,其应用并不总是可行的,且因为一些技术和经济原因,在可行的情况下,其应用也并不是简单直接的。总结了一种更一般化的不确定度评定和表示的随机模糊变量RFVs方法,系统评述了其关键技术与难点,通过实例表明,RFVs方法在非线性测量函数中传递不确定度具有简单高效的特点,最后对该领域的研究扩展提出了两点建议,给出了使用RFVs扩展贝叶斯定理的初步探讨结果。RFVs方法基于数学可能性理论,从GUM及其基本概念和定义出发对GUM方法进行了扩展,具有明显的优势,该方法的广泛应用也证明了其远大的发展前景。更多还原

【Abstract】 The"Guide to the Expression of Uncertainty in Measurement(GUM) "has been universally accepted and the recommendation of associating an uncertainty value with the measured value is universally adopted.However,this document follows an intrinsic probabilistic approach,its application is not always feasible;and when it is feasible,the application is also not straightforward due to some technical and economic reasons.In this paper,a more generalized random-fuzzy variables(RFVs) method for evaluating and expressing measurement uncertainty is briefly summarized and its key techniques and difficulties are systematically presented.An example is given to show that the RFVs approach is convenient and efficient when it is used to propagate uncertainty through nonlinear measurement function.Finally,two suggestions are proposed to extend the study in this field,and especially the preliminary investigation results for expanding Bayesian theorem with RFVs are given.The RFVs method is based on the mathematical theory of possibility,and expands the GUM approach starting from the GUM and its fundamental concepts and definitions;this method has obvious advantage,and the extensive applications of the method also show its great promising future.更多还原