摘要
采用近红外光谱分析技术,对干辣椒辣度的校正模型的影响因素进行比较研究。高效液相色谱测定干辣椒样品中的辣椒碱和二氢辣椒碱含量,计算得到辣度,并用近红外光谱仪采集试样的近红外漫反射光谱,研究样品粒度、光谱波段和预处理方法及主成分数选择对校正模型的影响,采用偏最小二乘法(PLS)建立干辣椒样品的辣度评价的校正模型,并对模型进行预测。结果表明,样品粒度为0.1～0.4mm,筛选出建模光谱波段为9000～4000cm-1,预处理方法为移动平均平滑、归一化处理和多元散射校正处理,最适主成分数为8;在优化条件下所建立的近红外光谱预测模型效果最好。
Near infrared reflectance spectroscopy （NIRS） is applied to research the influence factors of pungency of dried peppers. Quantitative analysis of capsaicin and dihydrocapsaicin is carried out using HPLC system. The influences of sample particle sizes, spectral waveband, pretreatment method and principal components on the calibration model by PLS are investigated. The results show that the sample particle sizes are 0. 1-0.4 mm, spectral waveband of 9000-4000 cm^-1 , pretreatment method is moving average, normalizing and multiple scattering correction, and principal components of 8. The best near infrared reflectance spectroscopy prediction model is built under the above conditions.
出处
《中国调味品》
CAS
北大核心
2013年第2期92-95,117,共5页
China Condiment
基金
重庆市自然科学基金(CSTC
2008BB9374)
关键词
近红外光谱
辣椒
辣度
影响因素
NIRS
dried peppers
pungency
influence factor
