摘要
探讨应用近红外光谱技术测定大豆油的过氧化值。通过比较各种数据处理方法,以143份样品建立大豆油过氧化值(0.5~50 mmol/kg)的近红外定标模型,70份样品作为验证样品集进行了验证。结果表明:以一阶导(DB1)结合0~1间归一化(N01)后采用偏最小二乘法(PLS)建模效果最好,定标集预测标准差(SEC)0.516,相关系数(Rc)0.999,验证集预测标准差(SEP)0.523,相关系数(Rv)0.998,表明利用近红外光谱技术能快速测定大豆油的过氧化值。
Peroxide values(0.5~50 mmol/kg) of soybean oil samples were determined by near-infrared(NIR) spectroscopy.Results show that the model with first derivative combined normalization between 0 and 1(N01) is the best.The standard error of calibration(SEC),standard error of validation(SEP),correlation coefficient(r) of calibration set(C-set),and validation set(V-set) are 0.5164,0.5230,0.9986,0.9980,respectively.The result demonstrates that NIR spectroscopy technology can be successfully applied as a rapid method for determining peroxide value(PV) of soybean oil.
出处
《中国粮油学报》
EI
CAS
CSCD
北大核心
2008年第6期206-209,共4页
Journal of the Chinese Cereals and Oils Association
基金
国家标准制定计划项目(20071625-T-449)
关键词
大豆油
近红外光谱
过氧化值
soybean oil,near-infrared spectroscopy,peroxide value