近红外漫反射光谱最小二乘法测定饲料中水份含量

Near-IR Reflectance Spectroscopy Analysis of Moisture in Forage by Least Square Regression

本文用经验波长法和相关系数(R)-斜率(K)波长优选法对鳗鱼饲料样品的近红外漫反射光谱和水份含量进行多元线性最小二乘回归分析,结果表明R-K法可以快速、无损地预测样品的水份含量,相对误差小于15％。

A multivariate linear least square regression analysis with experimental & R-K' s wavelenghts was used in the prediction of forage moisture by near infrared diffuse reflectance spectra. The R-K method can predict the moisture content of forage with convenience and nodestruc-tion, relative error<15%.