查詢結果分析
來源資料
頁籤選單縮合
題名 | FAST與KES-FB織物總手感換算公式之研究=Establishment of FAST and KES-FB Fobric Total Hand Value Translation Equation |
---|---|
作者 | 石天威; 林哲彥; 賴顯松; | 書刊名 | 紡織中心期刊 |
卷期 | 12:3 2002.07[民91.07] |
頁次 | 頁153-160 |
分類號 | 478.1 |
關鍵詞 | 總手感值; KES-FB系統; FAST系統; 迴歸; 類神經網路; Total hand value; KES-FB system; FAST system; Regression; Neural network; |
語文 | 中文(Chinese) |
中文摘要 | 由於FAST系統測得的某些織物物理性質與手感間存在相關性。為此,本研究透過統計方法,試圖建立織物物理性質與KES-FB總手感的換算合式。由實驗結果發現,確實可由織物物理性質去預測織物總手感。換算公式中的參數,係由在2gf/cm2 壓力下的平均厚度、在100gf/cm2壓力下的平均厚度、表面厚度、彎曲剛度、在20gf/cm荷重下之伸度、在100gf/cm荷重下之伸度、成型性及重量等八項所組成。類神經網路模式所架構的織物總手感預測模式,明顯優於迴歸法,其實際值與預測值間的相關係數分別為0.951及0.988。 |
英文摘要 | Certain physical properties measured by the FAST system are connected to the handle of the fabrics. For this purpose, this paper employed the statistics method in an attempt to establish the transformation model of the fabric physical properties and the total hand value of the KES-FB system. Experiment findings revealed that it is possible to predict the fabric total hand value through its physical properties. The parameters included in the transformation model are the thickness at 2gf/cm2 , the thickness at 100gf/cm2 , surface thickness, bending rigidity, extensibility at 20gf/cm, extensibility at 100gf/cm, formability, and weight of the fabrics. Model tests showed the presence of good prediction ability. It is very evident that the total handle value prediction model established by the neural network is more superior to the regression model. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。