Taguchi Method In The Assessment Of Measurement Uncertainty

The uncertainty of measurement is a parameter which associates with the measuring result. It represents the dispersity of reasonably measured value. The quality of measuring result depends on the uncertainty to a large extent. When measuring sample, the uncertainty of measurement and the measuring results should be given together. Therefore, the uncertainty of measurement is the most fundamental and characteristic index of the measuring system , and it is particularly important to the testing of the products quality. In traditional evaluation method, the uncertainty of measurement is composed of type A and type B uncertainty. Type A uncertainty is calculated through the standard deviation of repeatable experiments, while Type B uncertainty is estimated from experience. However, such method is difficult to work in practical terms as there are some limitations from the precision of the mathematical models and varieties of hypothesis during the evaluating process. Meanwhile, it is difficult to give the uncertainty of anyone within the measuring scale. The Taguchi method shows the quality of measurement with measurement error expense function and the S/N ratio of measurement speciality. This method can solve the problem to calculate anyone uncertainty. The principle and application of the uncertainty evaluating by traditional and the Taguchi method is introduced in this paper, and they are also validated by examples.