摘要
目的:利用近红外漫反射光谱检测网纹瓜可溶性固形物含量。方法:用偏最小二乘回归建模方法比较取样部位以及7种光谱预处理对建模的影响。结果:在有效信息波谱范围6101.9～5446.2cm-1和4601.5～4246.6cm-1,对3个取样部位(共9个检测点)的平均光谱应用偏最小二乘结合矢量归一化建立模型,其决定系数R2=0.8647,内部交叉验证均方根差RMSECV为0.412。用此模型预测验证集22个样本,模型的R2=0.8895,预测标准偏差RMSEP为0.339。结论:该模型能够满足网纹瓜可溶性固形物含量快速检测的要求。
The objective of the present research was to use NIR diffuse reflectance spectroscopy as a rapid and nondestructive method for the determination of soluble solids content of nutmeg melon.Statistical models were developed using partial least squre regression（PLSR） method by OPUS-QUANT2 software.The results of the PLSR models indicated that the models based on average spectra of nine measurements in three different parts of each nutmeg melon and corresponding soluble solids content obtained better results than those models based on the flesh of one part of each fruit（three measurements）.Seven spectral data preprocessing methods such as first derivative（FD）,second derivative（SD） and vector normalization（VN） were also discussed,the results showed that the best statistical model was developed using PLSR with respect to VN in the wave number range of 6 101.9～5 446.2 cm-1 and 4 601.5～4 246.6 cm-1,The correlation coefficient（R2） of the model was 0.8647,the root mean square error of cross validation（RMSECV） was 0.412.Using the model to validate 22 samples,the results showed that R2 is 0.8895 and the root mean square error of prediction（RMSEP） was 0.339.It is concluded that the prediction of the model can meet the requirement of rapid measuring of nutmeg melon＇s soluble solids content.
出处
《中国食品学报》
EI
CAS
CSCD
北大核心
2010年第4期272-277,共6页
Journal of Chinese Institute Of Food Science and Technology
关键词
近红外
网纹瓜
光谱预处理
可溶性固形物
检测
NIR
nutmeg melon
spectra preprocessing
soluble solids content
measurement
