臺灣博碩士論文加值系統

隨著高科技精密產業快速發展，以及對品質要求水準的提昇，使得各類量測儀器所需的精準度相對的也提高。因此，如何確保量測系統之檢測品質益發重要。國內雖已逐漸重視量測儀器之品保，卻仍著重於儀器的校正與追溯，對於量測系統評估之探討，則相當缺乏。職是之故，本研究的目的在於建構計量量測系統之評估模式。
為達此目的，首先根據線性、穩定性、偏差、重複性與重現性五個特性，建構四階段量測系統評估模式之架構，並依此模式架構，分別建立其評估方法與評估準則。在評估方法方面，除了建立評估流程步驟，並應用線性迴歸模式、變異數分析等統計手法，以評估此五個特性。
同時，在評估準則方面，對於偏差、穩定性與線性此三個特性，利用假設檢定之統計方法，以個別檢定其是否符合既定之要求水準。而針對重複性與重現性，則採用學術界以及產業界所廣為接受使用之評估準則。最後，以電子式游標卡尺作為實例驗證的主題，以說明本研究所建構評估模式之運作過程以及驗證模式之可行性。

Due to the rapid growth of high technology industry and higher standard for quality, the requirement for the higher accuracy of measurement instruments is expected. Despite the quality assurance of measurement instruments has caught the attention, but only focus on the traceability and calibration of instruments. It also lacks of discussions about the evaluation of measurement systems. So, the purpose of this research is to establish the evaluation model for the quantitative measurement systems.
For this purpose, measurement errors are divided into five categories: linearity, stability, bias, repeatability, and reproducibility. Firstly, based on these five characteristics, the four-stage evaluation model of the measurement system is constructed. The model developed also includes evaluation methods and criteria. For the evaluation methods, statistical methodologies such as linear regression analysis and ANOVA are applied.
Meanwhile, this research applies the statistical hypothesis to test linearity, stability, and bias whether these three characteristics meet the requirements or not. While considering the repeatability and reproducibility, this research uses the general evaluation criterion commonly accepted by the academic systems and the reality. Finally, the verification of this research is conducted by using the electronic caliper example.