摘要
用厚度为 1cm的玻璃比色皿作为吸收池 ,测定了亚麻油的近红外光谱 ,利用 70 0 0～ 6 0 0 0cm-1范围内的透波率 ,建立了主成分回归分析模型和BP人工神经网络模型 ,用二模型预测了亚麻油中油酸、亚油酸和亚麻酸的含量 ,预测结果的平均相对误差均在 2 %以内 ,并BP人工神经网络模型预测的效果较好。该法可用于亚麻油中主要组分的实时成分分析。
Using a 1cm glass cell as the absorption cell, the spectra of linseed oil was determined and analyzed. The transmittance in 7000～6000cm -1 were analyzed with principal component regression and BP artificial neural network (ANN), and the content of oleic acid , linoleic acid and linolenic acid in linseed oil were forecasted. The average of the relative error of the forecasted results was less than 2%, and BP ANN had a better outcome. This method can develop into an real time quantificational analysis of primary components linseed oil.
出处
《河南科学》
2002年第3期245-248,共4页
Henan Science
关键词
近红外光谱法
定量分析
亚麻油
组分
人工神经网络
主成分回归
linseed oil
near infrared spectroscopy
artificial neural network
principal component regression
