2023-06-142024-05-13https://scholars.lib.ntu.edu.tw/handle/123456789/653543茶葉的品質受到多種因素的影響，包括氣候條件、田間栽培的方式、製茶加工的過程，以及商業拼配的策略。因此，藉由統計分析進行茶葉相關試驗研究，以增進茶葉相關科學研究，政策執行與產業發展，本計畫將著重於以下四項研究： (一)建構影響茶葉品質因子資料分析：透過茶改場提供全國各示範茶園之茶樣修剪日，採收日與茶乾等基本資料、當地氣象數據及化學成分(如總氮、單寧、游離胺基酸、茶胺酸、兒茶素、咖啡因、維生素C…等)含量數據，透過數據分析，建立茶葉等級、氣候因子及化學成分含量的關聯性，並找出影響茶葉品質的關鍵成分含量及茶菁性狀標準。 (二)製茶工序參數資料分析：依茶改場收集之全國製茶技術競賽製茶資料，包括各參賽選手製茶過程中之萎凋、攪拌、發酵、炒菁、揉捻與乾燥等參數資料，及各參賽選手製茶品質(得獎名次)，經由數據處理與統計分析，找出茶葉品質較佳之製茶步驟與參數，以提供未來自動化製茶機械相關參數開發之用。 (三)商用茶及精品紅茶原料的AI拼配程式開發：根據已分級及風味輪評鑑之商用茶及精品紅茶原料數據，分析其化學成分後，建構成分與等級、風味之統計模型，進而開發AI拼配程式（自動化拼配系統），降低人為感官品評的誤差，讓產品穩定性更為提升。 (四)建置文山包種茶資料庫及優化烏龍茶和東方美人茶產地鑑別資料庫之統計分析：透過由茶改場提供文山包種茶、烏龍茶及東方美人茶茶樣之元素分析資料，藉由統計模式，建置文山包種茶資料庫並優化烏龍茶和東方美人茶產地鑑別模型。並建立境內境外不同混茶比例的鑑別模型，以利執行檢驗的實際需求。 以上的研究目標，將協助我們更深入了解影響茶葉品質的各項因素，並提出改善的策略和方法。 The quality of tea leaves is affected by various factors, including climatic conditions, field cultivation methods, tea processing procedures, and commercial blending strategies. Therefore, by conducting tea-related experimental research through statistical analysis, this project aims to enhance scientific research related to tea, policy implementation, and industry development. The project will focus on the following four research areas: (1) Constructing data analysis of factors affecting tea quality: Through data provided by tea improvement stations nationwide, such as tea sample pruning day, harvesting day, and dry tea basic data, local meteorological data, and chemical components (such as total nitrogen, tannins, free amino acids, catechins, polyphenols, caffeine, vitamin C, etc.), we will conduct data analysis to establish the correlation between tea grades, climatic factors, and chemical content, and find out the key component content and tea shoot traits that affect tea quality. (2) Data analysis of tea making process parameters: Based on the data collected from national tea making technical competitions held by tea improvement stations, including parameters such as withering, stirring, fermentation, pan-firing, rolling, and drying used by each competitor during the tea making process, and the quality of the tea they made (rankings), we will process and statistically analyze the data to determine the better tea making steps and parameters for future development of automated tea making machines. (3) Development of AI blending program for commercial tea and boutique black tea raw materials: By analyzing the graded and flavor-profiled commercial tea and boutique black tea raw materials, we will build statistical models of component, grade, and flavor relationships, and then develop an AI blending program (automated blending system) to reduce errors in human sensory evaluation and improve product stability. (4) Establishing the Pouchong tea database and optimizing statistical analysis of the Oolong and Oriental Beauty tea origin identification databases: Using elemental analysis data from Pouchong, Oolong, and Oriental Beauty tea samples provided by tea improvement stations, we will establish the Pouchong tea database and optimize the Oolong and Oriental Beauty tea origin identification model. We will also establish an identification model for different ratios of blended teas both domestically and internationally to meet the practical needs of inspection. The above research goals will help us understand more deeply the various factors that affect the quality of tea leaves and propose improvement strategies and methods.茶葉品質; AI 拼配; 製茶工序; 產地鑑別;;tea quality; AI blending program; tea making process; origin identification