摘要
用因子分析法解析了饲料样品的近红外漫射光谱,利用前面8个主成分因子与样品的化学分析结果拟合校正曲线,同时预测了未知样品的水分、粗蛋白和粗脂肪含量,该法无损、快速、准确,平均预测误差对水分为3.6%,粗蛋白为2.1%,粗脂肪为8.1%。
The near infrared diffuse reflectance spectra of forage can be analysed by pincipal factor analysis.The calibration curve fitting from frist eight principal factor scores and chemical analysis results can simultaneously predict the contents of crude protein,moisture,crude fat of unknown forage samples with nondestructive analysis,the average estimate error is 2.1%,3.6%,8.1% respectively for moisture,crude protein and crude fat.
出处
《福建分析测试》
CAS
1995年第3期301-306,共6页
Fujian Analysis & Testing