基于数据归一化原点矩约束最大熵原理的小样本数据测量不确定度评定

Measurement Uncertainty Evaluation of Small Batch Data Based on Maximum Entropy Principle Constrained by Data Normalized Origin Moment

测量检验是生产制造的重要环节,测量不确定度是测量质量的重要评价指标,表征测量数据的分散程度。最大熵原理常用于测量不确定度评定,基于最大熵原理获得测量样本数据的概率密度函数,能够避免主观因素,反映数据客观的分布情况。小批量生产中获得的测量数据往往较少,在小样本数据下,基于原点矩约束最大熵原理求得的测量不确定度不够准确可靠。针对这一问题,提出基于数据归一化原点矩约束最大熵原理的小样本数据测量不确定度评定方法,并通过数值仿真和应用案例验证所提测量不确定度评定方法的有效性。

Measurement and inspection is an important link in manufacturing,and measurement uncertainty is an important evaluation indicator of measurement quality,which characterizes the degree of dispersion of measurement data.The maximum entropy principle is often used to evaluate the measurement uncertainty.The probability density function of the measured sample data can be obtained based on the maximum entropy principle,which can avoid the subjective factor and reflect the objective distribution of data.The measurement data obtained in small batch production is often less,and the measurement uncertainty obtained based on the maximum entropy principle constrained by the origin moment is not accurate and reliable enough.To address this problem,a measurement uncertainty evaluation method for small batch data based on the maximum entropy principle constrained by data normalized origin moment was proposed,and the effectiveness of the proposed measurement uncertainty evaluation method was verified through numerical simulation and application case.