对测量仪器“示值重复性”评定方法的探讨

Research on the Method for Evaluating Indication Repeatability of Measuring Instrument

“示值重复性”是某些测量仪器的重要精度参数之一 ,它反映测量仪器在同一条件下的随机变化。对不同的测量仪器 ,这一参数具有完全相同的统计学含义。但目前尚未对示值重复性给出一个统一的规范化的评定方法 ,不同的测量仪器的检定规程或标准给出了不同的评定方法 ,因而所作的评定具有很大的随意性。为使示值重复性评定具有一致性和可比性 ,并在测量使用中引用这一参数合成不确定度时具有确定一致的可信度 ,应将示值重复性这一参数视为不确定度分量进行分析和估计 ,给出规范的统一的评定方法。根据统计学原理和有关法规文件 ,将示值重复性作为统计方法估计的 A类分量评定 ,即由测量仪器重复测量数据用 Beseel公式计算标准差 ,按 t分布给出扩展不确定度 ,并给出相应的置信概率和自由度。

Indication repeatability is one of the important precision parameters of some measuring instruments. It reflects the random changes of the indication of the measuring instrument in the same condition. This parameter has the same statistical meaning for different measuring instruments. However, there is no uniform and standard evaluating method yet. The regulations or standards of the verification of different measuring instruments present different evaluating methods. Therefore, the evaluation takes on great randomicity. To make indication repeatability possess consistency and comparability, and when introducing this parameter to synthesize uncertainty in the measurement have certain and consistent reliability, indication repeatability should be analyzed and estimated as uncertainty component, a standard and uniform evaluating method is given. According to statistics principles and the rules of laws and the documents involved, indication repeatability is accessed as type A component which is evaluated by statistic methods, that is, using measuring instruments measure data repeatedly, calculating standard deviation via Bessel formula, giving its expanded uncertainty in terms of t distribution, and presenting corresponding level of confidence and degree of freedom.