【摘要】：為快速檢測貯藏花生的質(zhì)量安全,對滅菌后的新鮮花生仁樣品分別接種5種常見的有害霉菌,并于26℃、相對濕度80%條件下貯藏9 d。利用衰減全反射-傅里葉變換紅外光譜(attenuated total reflectance-Fourier transform infrared spectroscopy,ATR-FTIR)采集不同貯藏階段花生樣品在4 000~600 cm~(-1)的光譜信息,通過權(quán)重分析闡述花生中侵染霉菌后光譜特征的變化,并結(jié)合偏最小二乘回歸(partial least squares regression,PLSR)分析建立樣品有害霉菌污染的定量分析模型。結(jié)果表明,不同貯藏階段樣品的峰譜出現(xiàn)明顯波動,PLSR模型對單一菌株與多種菌株樣品菌落總數(shù)的預(yù)測精度較高,其中對赭曲霉3.6486處理組樣品預(yù)測模型相對偏優(yōu),有效決定系數(shù)(R_p~2)為0.915 7、交互驗證均方根誤差(root mean-square error of cross-validation,RMSECV)為0.208 0(lg(CFU/g))、剩余預(yù)測偏差(residual predictive deviation,RPD)為2.52;對多種菌株預(yù)測結(jié)果R_p~2、RMSECV、RPD分別為0.780 3、0.358 0(lg(CFU/g))與1.76。應(yīng)用ATR-FTIR技術(shù)對花生受霉菌侵染的狀況進行快速分析具有可行性。
[Abstract]:In order to quickly detect the quality and safety of stored peanut, fresh peanut kernel samples after sterilization were inoculated with 5 kinds of common harmful fungi, and stored at 26 鈩，

本文編號：2274057