16
Views
0
CrossRef citations to date
0
Altmetric
Original Articles

應用模糊理論於等級類別變數之參數最佳化

Applying fuzzy set theory to achieving parameter optimization for the ordered categorical quality characteristic

&
Pages 425-435 | Published online: 30 Mar 2012
 

摘要

一般產品的品質特性通常可分成兩種型式:一種爲可量測的品質特性,另一種爲無法量測的品質特性。目前大多數的硏究都是針對可量測的品質特性做最佳化的探討,至於無法量測的品質特性由於在分析上較爲困難,因此在這方面的研究較少見。目前關於無法量測的品質特性之參數最佳化的一般作法,是先主觀地將這些無法量測之品質區分成數個類別(例如:以目視的方式來評估產品的均勻性時,可將產品品質的均勻性區分成均勻性好、均勻性尙可、均勻性不好等類別),然後再做分析。然而依據工程知識來主觀評估產品的品質時,可能會有模擬兩可(ambiguity)的情形,誓如:當缺陷嚴重的程度會影響到產品品質均勻性的評估時,則工程人員可能會因對缺陷嚴重程度認知的不同,導致其在品質均勻性主觀評估(subjective estimation)上造成差異,而使得這些無法量測的品質特性所隨隱含的資訊無法充分有效地定義出來。因此這些關於產品品質的人爲主觀的評估,在本質上大都具某種模糊性。此外,在製程參數最佳化方面,田口(Taguchi)的品質工程分析技巧可說是目前業界最常使用的製程參數最佳化方法之一,田口品質工程主要是結合實驗設計(experimental design)的特質與衡量品質損失函數(quality loss function)的觀點而成的分析技巧,以較少的實驗次數來達成有效的分析。至於在處理等級類別變數的參數最佳化方面,田口雖然也發展出一個累積分析法(accumulation analysis,或簡稱AA),但是此法無法有效地將主觀判定(例如:均勻性好、均勻性尙可、均勻性不好等)所能提供的資訊引入製程參數最佳化的分析中。其他的等級類別製程參數最佳化方法如分數設計法(scoring scheme,簡稱SS法)及加權機率分數設計法(weighted probability scoring scheme,簡稱爲WPSS法),也都無法適切地將工程知識或是工程經驗所隱含的資訊充份地引入製程參數最佳化的分析中,因此這三種方法所決定出來的最佳化參數設定都可能無法確實地達成參數最佳化之目標。由於模糊集合理論(fuzzy set theory)是一個對於語意性措辭相當有效的分析技巧,可以有效地將主觀的判斷予以數量化。因此,本研究將模糊集合理論引入田口的二次損失函數模式中,發展出—個摸糊損失函數(fuzzy-quality-loss-function,FQLF),以有效地衡量出利用主觀評估的方式所定義的等級分類型變數的品質損失,並以所計算之模糊損失函數値來進行等級分類型變數之製程參數的最佳化。本研究所提之方法可以將工程人員主觀判定的資訊引入製程最佳化的分析中,將可以增加傳統最佳化分析技巧的應用彈性。本研究最後引用一個半導體離子植入製程實例來驗證所發展之應用摸糊理論分析法於具有定性式等級類別變數之製程參數最佳化的有效性和 可行性。

Abstract

Quality characteristics of a product are frequently categorized as two forms: one is measurable characteristic and the other is non-measurable characteristic. The conventional way of optimizing a non-measurable characteristic is to divide the response of the non-measurable characteristic into several qualitative ordered categories according to the subjective judgement or estimation based on the engineering experience. However, the subjective judgement frequently causes some ambiguity and leads to different evaluation results based on the different subjective estimations. Fuzzy set theory is a well-known approach used to deal with the qualitative type or linguistic description of quality characteristic. The subjective estimations of the quality characteristic can be quantified well by employing the fuzzy set theory. Besides, Taguchi's quadratic quality loss function is a very useful index for evaluating the quality performance. In this study, the fuzzy set theory will be integrated into the Taguchi's quadratic loss function, and a fuzzy-quality-loss-function (FQLF) will be developed to achieve the parameter optimization for the qualitative ordered categorical quality characteristic. Different subjective estimations can be combined well by the proposed FQLF. An illustrative example, owing to the ion implanting process, will demonstrate the effectiveness of the proposed approach.

Reprints and Corporate Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

To request a reprint or corporate permissions for this article, please click on the relevant link below:

Academic Permissions

Please note: Selecting permissions does not provide access to the full text of the article, please see our help page How do I view content?

Obtain permissions instantly via Rightslink by clicking on the button below:

If you are unable to obtain permissions via Rightslink, please complete and submit this Permissions form. For more information, please visit our Permissions help page.