自适应蒙特卡洛法评定量块校准测量不确定度

Measurement uncertainty evaluation of gauge calibration using adaptive Monte Carlo method

测量不确定度表示指南(GUM)提供的不确定度评定方法,应用于模型非线性和输入量高相关的测量系统精度评价时存在一定的局限性。以GUM指南附录H.1包含了不确定度评定诸多细节的量块校准为对象,研究自适应蒙特卡洛法解决非线性模型和输入量强相关测量系统的不确定度评定难题。讨论了量块校准的测量模型和不确定度建模问题,给出了3种不确定度评定方法。基于蒙特卡洛计算机模拟的技术优势,以及自适应法仿真次数的优化,对量块校准测量不确定度评定进行了实例验证,结果表明,当测量模型为非线性、输入量强相关时,自适应蒙特卡洛法的计算效率最高。

The uncertainty assessment method provided by the guide to the expression of uncertainty in measurement(GUM)has significant limitations when the measurement model is nonlinear or input quantities are high correlation in measurement system accuracy evaluation.adaptive monte carlo method(AMCM)for solving uncertainty evaluation problem of nonlinear model and input quantities high correlation measurement system was studied based on the gauge calibration in the GUM guide appendix H.1.The measurement model and uncertainty modeling problem of gauge calibration were discussed,and three uncertainty assessment methods were given.Based on the technical advantages of Monte Carlo computer simulation and the optimization of the number of adaptive simulation,the measurement uncertainty of the gauge calibration was verified.And it can be concluded that AMCM is the most efficient when the measurement model is nonlinear and the inputs are high correlation.