(3.237.97.64) 您好!臺灣時間:2021/03/05 03:12
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:高敬先
研究生(外文):KAO- CHING-HSIEN
論文名稱:低鹽低糖蜜餞中微生物生長預測模式及其應用軟體之開發
論文名稱(外文):Predictive modelling of microbial growth and development of its application software on low-salt, low-sugar candied fruits.
指導教授:陳鴻章陳鴻章引用關係
指導教授(外文):Hung-Chang Chen
學位類別:碩士
校院名稱:大葉大學
系所名稱:食品工程研究所
學門:農業科學學門
學類:食品科學類
論文種類:學術論文
論文出版年:1998
畢業學年度:86
語文別:中文
中文關鍵詞:低鹽低糖蜜餞多項式迴歸分析數學模式預測微生物學
相關次數:
  • 被引用被引用:0
  • 點閱點閱:274
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:0
本研究探討加工條件,如鹽、糖濃度、防腐劑用量、儲藏溫度與包材透氣性(PE 袋、KOP CPP 袋、夾口袋)等與蜜餞品質及安全之關係,進而建立可靠之資料庫及數學預測模式,並將所得之數學預測模式利用 MicroSoft Visual Basic 5 程式開發語言轉化為對蜜餞業者及非研究人員具親和性之應用套裝軟體,以供業者使用,並助其加速開發新式低鹽低糖蜜餞產品及確保衛生與安全。 本研究首先蒐集金柑蜜餞所須之微生物品質資料庫 (總菌數及酵母菌與黴菌數共各蒐集到 27 × 9 × 11 數據點 ),並以甘培茲函數進行配適及統計分析,結果迴歸效果並不理想 (R^2 絕大部份小於 0.2 ),與前人迥異。此意謂甘培茲方程式可能只適合於液體培養基中短時間快速生長且無死滅的配適,而不適於固體基質中長時間儲藏之菌相生死變化的配適,而究其原因可能係因微生物在不同環境下之生長反應不同所致。 由於甘培茲函數無法正確地描述蜜餞中微生物生長情形,故將金柑微生物品質資料庫以總生菌數對加工及儲藏條件進行二次多項式直接迴歸分析,求取單一類包裝及整體包裝金柑蜜餞產品之總生菌數對糖、己二烯酸、水活性、儲藏溫度及儲藏時間等各因子間的迴歸方程式,其迴歸結果R^2 亦皆非常低,表示此一模式之契合效果仍然不佳。 最後,改以總生菌數經由生長參數對加工及儲藏條件進行二次多項式間接迴歸分析。典型之微生物生長曲線分成三個階段,包括(1)遲滯期(2)對數期及(3)穩定期 。本研究先以固定加工條件下之總生菌數變化對時間進行線性迴歸,求得單一加工條件下之參數,其分別為遲滯時間(t1)、比生長速率(a 或 m)、進入穩定期之時間(t2)及擬起始菌數(Y0),令其迴歸之相關係數為 R1^2;為使模式更能準確描述微生物生長情形,進一步將上述迴歸結果中 R1^2 大於 0.7 以上之各生長曲線參數利用多項式二次迴歸求得其與加工條件間之關係,令其迴歸係數為 R 2^2。R 1^2 代表生長曲線與生長模式之契合度;R 2^2 代表生長動力學參數 t1、t2、μ、α和 Y0 與加工條件之契合度。一個模式是否得到正確的配適決定於 R^2 的大小,R^2 越接近於 1 表示契合度越高;由於R1^2、R2^2是同等的重要,因此本研究選取 ( R1^2 + R2^2 )max狀態下之方程式為最適模式。此模式在對數生長時期準確度高達 85-95 %,對描述蜜餞中微生物生長情形而言是一個不錯的選擇。 目前的作業系統中 Microsoft Windows 95 或是即將上市的 Microsoft Windows98 都是一個極具親和力的使用界面,也是未來的趨勢,本論文中所發展之低鹽低糖蜜餞安全儲架期預測軟體是經過 Microsoft Visual Basic 5 封包的單一軟體,可直接應用於 Microsoft Windows 95 系統上,不須配合其他軟體就可以使用。除可在特定加工及儲藏條件下預測蜜餞產品之儲架期及特定儲架期下預測產品之單項規格外,該軟體並具有繪圖功能,可使使用者在極短的時間內就可看出菌數生長的變化,算是非常具有親和性的軟體。

This research investigated the effect of processing conditions, such asconcentration of salt and sugar, amount of preservatives, storagetemperature and oxygen permeability of the packaging material (PE bag, KOP/CPP bag, zipper bag )on the safety and quality of candied fruits. Suitabledatabases and mathematically predictive models were built from the datacollected. Computer programming language such as Microsoft Visual Basic 5was then used to shift the mathematically predictive models into applicationsoftware, that is user friendly to the manufacturer and non-researchers ofcandied fruits, to help them speed up the development of a better version oflow-salt-low-sugar candied fruits and to ensure sanitation and safety of theproducts. The first step proceeded with curve fitting and statistical analysis ofthe microbial quality database collected from candied kumquat processing (the data points for total plate count and yeast and mould count are 27x9x11,respectively )with Gompertz function. However, the regression result of suchcurve fitting was not encouraging (R^2 mostly lesser than 0.2 ). It meansthat Gompertz function may be only suitable for short-term fit of fastgrowing and non-deceasing cells in liquid medium, but not suitable for thelong-term storage fit of life-and-death cycling of microbes in solid medium.The reason may be the fact that microbes response differently in variousliving conditions. Due to the fact that Gompertz function doesn't accurately describe themicrobial growth status of candied fruits, the total viable plate count inthe microbial quality database of kumquat was directly fitted to the processand storage conditions using quadratic polynomial regression. The regressionequations of total plate count to sugar concentration, sorbic acidconcentration, water activity, storage temperature, and storage time wereobtained for candied kumquat packed in single and all packaging materials.The regression results as of R2 were also very low, meaning that the fittingresults of the models were still not good enough. Finally, the research thus proceeded with quadratic polynomialregression of total viable cell count to process and storage conditionsindirectly via growth parameters. The three stages of typical microbialgrowth curve include lag phase, log phase and stationary phase. Initially,the research proceeded with linear regression of total viable cell countunder fixed process condition to storage time to obtain the parameters oflag phase time (t1 )against, specific growth rate or deceasing rate (μ orα ), time reaching stationary phase (t2 ), and pseudo initial cell count (Y0 ).The relative regression coefficient was defined as R1^2.Then, thequadratic polynomial regression analysis was used to correlate theprocessing conditions with growth parameters which had R1^2 value greaterthan 0.7 to build a more accurately predictive modeling on microbial growth.The relative regression coefficent in such case was defined as R2^2.R1^2is the fitness of growth kinetics parameters (t1、t2、μ、α and Y0 )toprocess conditions. The accuracy of any model is determined by the readingof R^2 and the fitness gets higher when R^2 reading gets closer to 1. Thefinal equation obtained from (R1^2 + R2^2 )max was chosen as the mostsuitable model, because both R1^2 & R2^2 are equally important. The fitnessof the model is 85~95 % accurate in log phase, proving to be a good choicefor the description of microbial growth in candied fruits. For the moment, operation systems that are available in the market suchas Microsoft Windows 95 and the pending version of Microsoft Windows 98 areall very user-friendly interfaces. The application software, developed inthis research for the prediction of microbial quality of candied fruits, isa single software written by Microsoft Visual Basic 5, and is compatible toMicrosoft Windows 95 without the need of other software. In addition topossessing the capability of predicting the shelf life of candied fruitsprocessed and stored under fixed conditions and the capability of predictingthe value of single process or storage parameter under specified shelf life,the software also supports graphics function, which enables user to foreseethe microbial growth in a short period of time.

QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
系統版面圖檔 系統版面圖檔