自适应MCM测量不确定度软件研发

Development of adaptive Monte Carlo method software for the evaluation of uncertainty measurement

蒙特卡洛法(MCM)测量不确定度评定中模型复杂、计算量大,现有软件往往只在某一模型中适用,不同模型间相互独立,且评定结果缺少自适应过程.基于LabVIEW软件生成输入量X的伪随机数,对X概率密度函数(PDF)离散抽样,得到输出量Y的离散抽样值,进而设计了自适应MCM测量不确定度软件.本软件实现了常用模型数学公式的自定义,增强了软件的适用性,同时重点介绍了自适应增加样本量M的算法.通过JJF1059.2-2012规范中实例的计算,验证了软件在常用模型评定中的有效性.

The uncertainty evaluation of Monte Carlo method（MCM）measurement is complex with its large amount of calculation.The present software is usally used only in a certain model.The models are independent and the result of the evaluation lacks the adaptive process.On LabVIEW software,we designed the software for the adaptive MCM measurement of uncertainty.The pseudo random number input X was generated.The discrete sampling of the probability density function（PDF）of X was conducted to get the discrete sampling values of output Y.The software was used to self-define the general model of mathematical formula with strong self-adaption.The algorithm of adaptively increasing sample size M was focused.The effectiveness of the software in the evaluation of the general model was verified with the JJF1059.2-2012 example.