本文選題：均勻設(shè)計 + 主成分分析　； 參考：《食品科學(xué)》2017年04期

【摘要】：為優(yōu)化甘薯薯片油炸工藝,采用單因素和均勻設(shè)計相結(jié)合進行試驗,對均勻設(shè)計試驗數(shù)據(jù)進行多元回歸分析。單因素試驗表明:切片厚度、預(yù)干燥階段和油炸階段對含油量、L*值和b*值有較大影響;均勻設(shè)計逐步回歸分析表明:油炸時間對含油量影響顯著(P=0.011),食鹽水質(zhì)量分?jǐn)?shù)對含水量影響顯著(P=0.022),油炸溫度和油炸時間對L*值影響極顯著(P=0.001),預(yù)干燥時間和油炸時間對b*值影響顯著,食鹽水浸泡時間、食鹽水質(zhì)量分?jǐn)?shù)、預(yù)干燥溫度、預(yù)干燥時間和切片厚度對脆度的影響符合二次多項式回歸模型(P=0.001);指標(biāo)主成分分析表明:提取2個主成分能解釋87.4%的指標(biāo)信息,達到指標(biāo)降維目的;嶺回歸分析建立了綜合得分回歸模型,相關(guān)系數(shù)R為0.997,能很好地擬合產(chǎn)品的綜合得分;偏最小二乘法回歸分析預(yù)測最佳綜合評分工藝參數(shù)為切片厚度2 mm、熱燙時間1 min、質(zhì)量分?jǐn)?shù)1%食鹽水浸泡20 min、預(yù)干燥溫度60℃、預(yù)干燥時間70 min、油炸溫度150℃、油炸時間1 min,驗證綜合評分為0.89,綜合評分高于均勻設(shè)計試驗組最高值0.86。優(yōu)化工藝所得產(chǎn)品具有含油量低、顏色和脆度俱佳等特點,相關(guān)模型具有良好的預(yù)測能力。
[Abstract]:In order to optimize the fried technology of sweet potato chips, the single factor and uniform design were used to carry out the experiment, and the data of uniform design test were analyzed by multivariate regression analysis. Single factor test showed that slice thickness, pre-drying stage and frying stage had great influence on oil content L * and b *; The results of stepwise regression analysis of uniform design showed that the effect of frying time on oil content was significant, salt water content on water content was significant, frying temperature and frying time had very significant effect on L ~ * value, pre-drying time and frying time were very significant. The effect on the value of b * is significant, The effects of salt water immersion time, salt water mass fraction, pre-drying temperature, pre-drying time and slice thickness on brittleness were in accordance with the quadratic polynomial regression model (P0. 001), and the principal component analysis showed that the extraction of two principal components could explain 87.4% of the index information. The regression model of comprehensive score was established by ridge regression analysis, and the correlation coefficient R was 0.997, which could fit the comprehensive score of the product well. The optimum synthetic scoring parameters were predicted by partial least square regression analysis: slice thickness 2 mm, blanching time 1 min, soaking 20 min in 1% salt water, predrying temperature 60 鈩，

本文編號：1894374