微型近红外光谱仪检测水果可溶性固形物

Detection of the soluble solids content(SSC)in fruit using miniature NIR spectrometry

以滨松公司生产的C11708MA微型光谱仪为基础,在自行搭建的两套光谱采集平台上检测水蜜桃、梨子的可溶性固形物含量。采用了多种光谱预处理方法,结合PLS和LS-SVM建立水蜜桃、梨子可溶性固形物模型。实验结果表明,水蜜桃光谱经过标准化预处理,建立的LS-SVM模型效果最好,校正相关系数(Rp)和均方根误差(RMSEP)分别为0.8902和0.7703。梨子光谱经过CARS筛选得到46个变量,建立的PLS模型效果最好,Rp和RMSEP分别为0.7597和0.5783。验证了该光谱仪在水果可溶性固形物含量检测方面的应用的可行性,为进一步构建便携式水果可溶性固形物检测设备奠定了基础。

Two sets of spectral acquisition platform based on miniature NIR spectrometry Cll708MA were builded. Different spectral pretreatment methods were developed to establish PLS and LS-SVM regression model of SSC in peaches and pears. The Rp and RMSEP were 0. 8902 and 0. 7703 for peaches. The Rp and RMSEP were 0. 7597 and 0. 5783 for pears. The above results demonstrate that the method is feasible, and the data are useful for the further applications of miniature NIR spectrometry.