基于FFT的湿度传感器测量不确定度评定方法

Evaluation method of measurement uncertainty of humidity sensor based on FFT

目前气象计量领域测量不确定度评定一般采用GUM(Guide to the expression of Uncertainty in Measurement,测量不确定度表示指南)方法,该方法存在置信因子近似估计的问题。为进一步提升测量不确定度评定结果的可靠性,提出了一种基于FFT的卷积评定方法。利用FFT算法得到多个不确定度分量的合成分布,并根据指定置信水平准确求出合成分布的置信因子,进而确定扩展不确定度。利用该方法对一款湿度传感器测量不确定度进行了评定,并与GUM方法评定结果进行对比分析。结果表明,该方法计算得到湿度传感器的扩展不确定度为1.54%RH,比GUM方法评定结果(1.68%RH)更加精确,且能直观显示合成分布情况,有助于提高相关业务人员对测量不确定度的理解水平。

At present,the measurement uncertainty evaluation in the field of meteorological metrology generally adopts the GUM(Guide to the expression of Uncertainty in Measurement)method,and there is a problem that the confidence factor is estimated approximately.In order to further improve the reliability of measurement uncertainty evaluation results,a convolution evaluation method based on FFT is proposed.The composite distribution of multiple uncertainty components of the sensor is calculated by using the FFT algorithm,and the confidence factor of the composite distribution is accurately obtained according to the specified confidence level,and then the expanded uncertainty is determined.This method is used to evaluate the measurement uncertainty of a humidity sensor,and the comparison is made with the evaluation results of the GUM method.The results show that the expanded uncertainty of the humidity sensor calculated by this method is 1.54%RH,which is more accurate than the evaluation result of the GUM method(1.68%RH),also the method can display the distribution directly,which is helpful to improve the understanding of the measurement uncertainty for the relevant business personnel.