论文部分内容阅读
[目的]探索一种快速测定完整藜麦籽粒纤维含量的方法。[方法]采集100个藜麦样品的近红外光谱,运用近红外光谱分析技术建立数学模型并进行预测。[结果]在10 000~4 000cm~(-1)波长范围内,运用一阶导数+矢量归一化光谱方法进行预处理,结合化学方法所得数据建立藜麦粗纤维近红外定量模型,校正和预测效果最佳,所得的粗纤维近红外定量模型的交叉验证决定系数(R~2cv)为0.884 8,外部验证决定系数(R~2val)为0.876 1。[结论]以完整藜麦籽粒为样品所建立的纤维NITS模型可用于藜麦纤维含量的快速检测。
[Objective] The research aimed to explore a method to quickly determine the fiber content of whole quinoa kernels. [Method] Near infrared spectra of 100 quinoa samples were collected and mathematically modeled by near infrared spectroscopy. [Result] The first derivative and vector normalized spectral method were used for pretreatment in the wavelength range from 10 000 to 4 000 cm -1. The quantitative model of quinoa crude fiber near infrared was established based on the data of chemical methods. The calibration and The prediction results are the best. The cross validation coefficient (R ~ 2cv) of the crude fiber near infrared quantification model obtained is 0.884 8, and the external validation coefficient (R ~ 2val) is 0.876 1. [Conclusion] The NITS model based on intact quinoa grains could be used for rapid detection of quinoa fiber content.