MUKOM - Cost-efficient determination of the measurement uncertainty of complex measurement processes

每次测量的结果都受到不确定度的影响。由于这种测量的不确定性，符合性检查可能导致有两个后果的错误测试决策。要么报废好零件，要么将故障零件交付给客户。为了确保测试决策的正确性，有必要了解测量过程的不确定度。现在，利用现有的程序，已经可以确定大量测量过程的测量不确定度。然而，这仍然是测量过程的一个挑战，因为没有适用的标准化计量模型。这些可以概括为“复杂测量过程”。例如硬度测试、具有大量影响量的检验或材料不均匀性的检验。然而，在研究和行业中，对这种程序有可验证的需求。因此，本研究项目的目标是为复杂的检测过程开发一种测量不确定度的估计程序。该程序使用户能够识别相关的影响量，模拟复杂的检查过程，并将专家的知识纳入建模。与传统方法相比，新开发的方法将采用一种创新的方法，根据科学接受的GUM（测量不确定度表达指南）进行分析建模，并从实验数据中推导元模型。因此，人们可以受益于基于先前知识的精确系统建模和使用系统测试的实验建模的优点。

The result of every measurement is impaired with an uncertainty. Because of this measurement uncertainty, the inspection of conformity can result in a faulty test decision with two consequences. Either good parts are scrapped or malfunctioning parts are delivered to the customers. In order to make sure that the test decision is correct, it is necessary to know the uncertainty of the measurement process. Right now and with the currently available procedures, it is already possible to determine the measurement uncertainty for a large amount of measurement processes. However, this remains a challenge for measurement processes, for which there are no standardized metrological models applicable. These can be summarized under the term complex measurement processes. Examples are hardness testing, inspections with large amounts of influencing quantities or inspections with inhomogeneities in the material. Nevertheless, there is a verifiable need in research and the industry for such a procedure. The goal of this research project is therefor to develop a procedure for the estimation of the measurement uncertainty for complex inspection processes. The procedure qualifies the user to identify relevant influence quantities, model complex inspection processes, include knowledge from experts into the modeling. Compared to the conventional procedures, the newly developed procedure will use an innovative approach, which combines the analytical modeling in accordance to the scientifically accepted GUM (Guide to the Expression of Uncertainty in Measurement) and the deduction of meta models from experimental data. Due to this, one can benefit from the advantages of the exact system modeling based on previous knowledge and the experimental modeling using systematic tests.