摘要
为实现玉米皮渣发酵生产微生物油脂过程中菌丝体油脂含量的实时监测,建立了菌丝体油脂含量测定的近红外定标模型。结果表明,采用偏最小二乘法(PLS)、归一化预处理方法,模型的定标集相关系数(R)及验证集相关系数分别为0.997 4和0.995 8,定标集标准偏差(SEE)为1.266 6,验证集标准偏差(SEP)为1.417 1,定标集标准偏差(SEE)与验证集标准偏差(SEP)的比值为0.893 7,这表明该方法可快速准确地测定菌丝体油脂含量。
The near - infrared spectrum model for determination of microbial oil content in mycelium was set up so as to realize the real - time detection of the oil content in mycelium during the microbial oil production from corn fiber by fermentation with Mortierella alpina. The partial least square regression（PLS） and the spectral data pretreatment of normalization were used to analyze the model. The results showed that the correlation coefficient （R） of calibration set （ C - set） and validation set （ V - set） were O. 997 4 and 0. 995 8, respectively, standard error of calibration set （SEE） was 1. 266 6, standard error of cross - validation （SEP） was 1. 417 1 ,the ratio of SEE to SEP was 0. 893 7. The method could measure the oil content in mycelium rapidly and accurately.
出处
《中国油脂》
CAS
CSCD
北大核心
2011年第1期77-79,共3页
China Oils and Fats
基金
国家自然科学基金(50673089)
吉林省教育厅重点项目基金资助(20081002009121)
作者简介
潘安龙（1984），男，硕士研究生，有机化学专业（E—mail）anlong107@yahoo．com．cn。
通信作者：薛冬桦，教授，博士（E—mail）xuedonghua@mail．ccut．edu．cn。
