近红外漫反射光谱法快速预测蒸汽压片玉米成分的研究

[Research on using near infrared diffuse reflectance spectroscopy for rapid prediction of compositions of steamed-flaking corn].

作者信息

Xue Feng, Wang Li, Meng Qing-Xiang, Cui Zhen-Liang, Ren Li-Ping

机构信息

State Key Laboratory of Animal Nutrition, China Agricultural University, Beijing 100193, China.

出版信息

Guang Pu Xue Yu Guang Pu Fen Xi. 2011 Jan;31(1):62-4.

PMID:21428057

Abstract

Near infrared diffuse reflectance spectroscopy (NIDRS) calibrations of chemical composition in 62 kinds of steamed-flaking corn were developed by partial least squares (PLS) regression. The calibration models were constructed by the mean centre + first deviation + multivariate scattering calibration in the square regions (950-1 650 nm). Results showed that the R2(VaL) are 0.951 1, 0.903 2, 0.714 3 and 0.908 2 for CP, NDF, ADF and EE respectively. The SECV are 1.53, 1.89, 2.45 and 0.50 for CP, NDF, ADF and EE respectively. Therefore, the models could be used to predict the content of CP, NDF and EE in steamed-flaking corn and provide a fast and economical method for quality control of steamed-flaking corn.

摘要

采用偏最小二乘法（PLS）回归建立了62种蒸汽压片玉米化学成分的近红外漫反射光谱（NIDRS）校正模型。在950 - 1650 nm的方形区域内，通过均值中心化 + 一阶导数 + 多元散射校正构建校正模型。结果表明，粗蛋白（CP）、中性洗涤纤维（NDF）、酸性洗涤纤维（ADF）和粗脂肪（EE）的交叉验证决定系数（R2(VaL)）分别为0.951 1、0.903 2、0.714 3和0.908 2。CP、NDF、ADF和EE的交叉验证标准误（SECV）分别为1.53、1.89、2.45和0.50。因此，该模型可用于预测蒸汽压片玉米中CP、NDF和EE的含量，为蒸汽压片玉米的质量控制提供一种快速且经济的方法。