方便面油分近红外漫反射光谱特征波长的确定

建立反映方便面含油率与近红外漫反射光谱（ＮＩＲＤＲＳ）数据之间关系的数学模型即定标方程，是利用ＮＩＲＤＲＳ技术检测方便面含油率的前提，其中定标方程选入的特征波长与方便面含油率关系最密切，因此特征波长的确定是建立定标方程的关键本研究采用４０个方便面定标样品，其ＮＩＲＤＲＳ数据分别经一阶和二阶导数变换处理，运用逐步回归分析方法建立了相应的两个定标方程，复相关系数分别为０．９７８、０．９８６，标准差分别为０．４２４、０．３４２判定５７６９ｃｍ－１、５７８１ｃｍ－１两个波长是方便面油分在近红外区的特征吸收波长为证实特征波长的可靠性，研究通过实验和理论分析比较了棕榈油、面粉—棕榈油、方便面、面粉和水分的近红外漫反射光谱的吸收特征，证明用逐步回归方法挑选出的如上两个特征波长是可靠的

The Determination of NIRDRS Characteristic Wavelength of Oil in Instant Noodles

In order to apply NIRDRS to oil measurement in instant noodles, calibration equation (CE) must be acquired firstly. It's the key to determine the characteristic wavelength which is closely related to spectral absorption of oil.Forty samples were used for instrument calibration by stepwise regression (SW) and two calibrations equation were obtained with the following statistical values: multiple correlation coefficient (MCC)=0.978 and standard error of calibration (SEC)=0.424 for first derivative spectroscopy-data-based CE, and MCC=0.986 and SEC=0.342 for second derivative spectroscopy-data-based CE. The wavelengths of 5769cm -1 and 5781cm -1 were recognized as characteristic wavelength of oil in accordance with correlation coefficient relevant to the wavelengths selected in the calibration equations. The results of experimental and theoretical analysis given afterwards show that, 5769cm -1 and 5781cm -1 characteristic wavelengths are reliable and the wavelengths selected explain the two calibration equations well.