摘要
GUM法是评定测量不确定度的一般方法,以传播不确定度的方式提供输出量的包含区间,在线性模型条件下能够提供较为准确的评定结果,气象仪器检定结果的不确定度评定也采用GUM法进行。然而,当测量模型和计算较为复杂时,GUM法得到的结果可能出现较大的偏差。蒙特卡洛方法(MCM)以传播概率分布的方式,为测量不确定度的评定提供了一种通用的数值方法。以气压传感器检定结果为例,运用MCM进行不确定度评定,并用MCM评定结果验证了GUM法评定结果的准确性。结果表明,GUM法评定结果未能通过验证,说明GUM法在气压传感器检定结果的不确定度评定方面可能存在一定的问题,而用MCM可以得到更为准确可靠的包含区间。
GUM is a common method for uncertainty evaluation.Inclusion intervals are calculated by propagating uncertainty with GUM,which can get accurate evaluation results.Uncertainties of meteorological instruments verification results are also evaluated by GUM.However,GUM results may appear deviation when measurement model and calculations are complex.Monte Carlo Method（MCM）provides a general numerical method for uncertainty evaluations by propagating probability distribution.Take verification results of air pressure sensors as examples,uncertainty was evaluated by MCM and GUM results were verificated.Results showed that GUM results did not pass the verification.There are some problems in evaluating air pressure uncertainty with GUM.A more accurate inclusion interval can be calculated by MCM.
出处
《国外电子测量技术》
2016年第8期90-93,98,共5页
Foreign Electronic Measurement Technology
