基于蒙特卡罗仿真的校准结论风险分析

Calibration conclusion risk analysis based on Monte Carlo simulation

针对已有的单次校准结论风险,计量实验室更加需要总体校准结论风险来分析测量仪器的校准状态,提出将蒙特卡罗仿真方法应用于校准结论的风险分析中。通过对单次校准结论的风险分析得出影响误判率的参数;进而通过蒙特卡罗仿真法分析总体校准结论,结果显示,提高测量不确定度比一倍,误判率下降120%。最后,结合仿真过程建立总体校准风险的模型,通过分析总体校准结论的风险,提出可以降低错误判断风险的概率的方法,对校准工作具有一定的理论指导价值。

Aiming at the existing single calibration conclusion risk, the analysis of overall calibration conclusion risk is more needed to assess the status of measuring instruments in metrology laboratory. In this paper, Monte Carlo simulation method is used for analyzing the calibration conclusion risk. The parameters that have effect on the false rate are gotten by analyzing the single calibration conclusion risk, and Monte Carlo simulation method is adopted to analyze the overall conclusion. The result shows that the misjudgment probability declining 120% when doubling uncertainty ratio. Finally, an approach is presented to reduce the misjudgment probability through overall calibration conclusion risk analysis combined with the model built in the simulation process, which has some guidance to the calibration work.