應用模糊多準則決策驗證咖啡豆最佳烘焙程度

Application of Fuzzy MCDM to Assess the Best Degree of Roasting of Coffee Beans

卓弘為(Hung-Wei Cho)

指導教授 ： 游文龍

國立虎尾科技大學/管理學院/工業工程與管理研究所/碩士(2014年)

https://doi.org/10.6827/NFU.2014.00076

摘要

咖啡豆的烘焙程序除了將豆體炒熟外，尚須藉由高溫焙炒過程催發咖啡豆產生風味與香氣。咖啡豆除了其產地與沖泡方式等要素會影響風味外，最大的關鍵則在生豆的烘焙加溫過程，藉由高溫催化豆中化學物質產生多種繁複的化學及物理反應，產生多種香氣分子，且藉著不同程度的溫度調控，即使是同種咖啡豆，沖煮出的咖啡風味表現表現也大大不同。
本研究目的在建構一種可判斷出咖啡特徵表現的評鑑模式，以及探討專家們主要多以那些因素作為咖啡優劣表現的評估重點，研究內容設計三種不同烘焙的加溫程序樣組，針對確立之烘焙模式進行實際烘焙並以電子化溫度量測模組進行溫度、時間等參數紀錄蒐集，將實際烘焙樣品分為「淺焙」組、「中焙」組與「深焙」組供後續評估嚐測。評估方式使用模糊德爾菲法（Fuzzy Delphi Method）評選咖啡的關鍵評鑑因素，並利用模糊層級分析法（Fuzzy Analytic Hierarchy Process）探討各因素間的重要程度，樣本的評選則使用模糊綜效評判進行績效評測，最後獲得之模糊績效結合模糊層級分析法中因素重要程度之模糊權重，進行加權運算與解模糊化，以獲得最佳的咖啡豆烘焙程度候選樣組。
實驗結果為「深焙」之樣組獲得最高的評價，表示本研究在「深焙」的加溫烘焙過程在各項評鑑因素表現上，經過模糊加權後較其它兩組優秀，另外於模糊層級分析中「順口度」、「酸表現」和「滋味變化」在重要程度上為排序前三高，顯示了專家們在感受或評價咖啡時的首要著重因素，可供相關組織或商家甚至個人在烘焙、評鑑咖啡時的依據，以及作為往後相關研究或調查的研究參考。

關鍵字

咖啡 ； 溫度量測 ； 評鑑機制 ； 模糊德爾菲法 ； 模糊層級分析法 ； 模糊綜效評判

並列摘要

The coffee bean roasting process not only involves frying beans, but also involves catalyzing the flavor and aroma of coffee beans through the roasting process under high temperatures. Other than “country of origin” and “brewing method”, the key lies in the bean roasting and heating process. Through varying degrees of temperature control, even the same kind of beans will produce greatly varied flavors after brewing.
The purpose of this study was to construct an evaluation model for determining the coffee characteristic performance and explore which factors experts use as the focus of coffee performance evaluation. The research content covers three different types of candidate heating procedural combinations during roasting. According to established roasting model, actual roasting was performed. Through the electronic temperature measurement module, parameters such as temperature and time were collected. The actual roast samples were divided into “light roast”, “medium roast”, and “dark roast” groups for follow-up assessment and taste testing. The fuzzy delphi method’s (FDM) key evaluation factors for coffee selection and the fuzzy analytic hierarchy process (FAHP) were adopted to explore the factor’s degree of importance. The fuzzy synthetic decision making was used for performance evaluation and testing. Finally, the fuzzy performance coupled with the fuzzy weights of the factors’ degree of importance, FAHP were used for weighted calculations and defuzzification, thereby obtaining the candidate sample group with the best degree of coffee bean roasting.
After performance measurement of the three sample sets, “dark roasting” received the highest rating. Additionally, in terms of degree of importance, FAHP analysis results show that “smoothness”, “Acidy”, and “Flavor” rank the top three, thus indicating the factors the experts will focus on when experiencing the taste of coffee. The findings shall serve as a basis for relevant organizations, stores, or individuals when roasting and evaluating coffee; they shall also serve as a research reference for follow-up research or survey.

並列關鍵字

Coffee ； Temperature-recording ； Assessment system ； Fuzzy Delphi Method ； Fuzzy Analytic Hierarchy Process ； Fuzzy synthetic decision making

參考文獻

18. 詹豐銘，2007，生咖啡豆在不同烘焙條件下焙炒後所調製咖啡之成分分析，國立臺灣大學，碩士論文。

12. 陳科后，2011，修枝對台灣阿拉比卡種咖啡樹之農藝性狀影響，國立臺灣大學，碩士論文。

4. 李宛凌，2008，台灣咖啡產業的歷史考察，國立臺北教育大學，碩士論文。

16. 葉協隆，2009，兩岸海運直航貨櫃港口選擇之研究，國立交通大學，碩士論文。

21. 蔡明諺，2013，應用良導絡經絡能量分析咖啡飲用後自律神經系統之即時反應，中國醫藥大學，碩士論文。