近红外光谱定量分析技术在枇杷可溶性固形物无损检测中的应用

Application of NIR Quantitative Analysis to Nondestructive Detection of Loquat Soluble Solid Content

应用近红外漫反射光谱定量分析技术对两个产地三个品种枇杷的可溶性固形物进行无损检测试验研究。通过分析，发现在波长1400～1500nm和1900～2000nm两段范围，样品的可溶性固形物与光谱吸光度之间的相关系数较高；用偏最小二乘回归PLSR、逐步多元线性回归SMLR和主成分回归PCR三种方法分别建立这两个波段和全波段范围的模型，全波段的PLSR模型的效果较优。研究发现一阶和二阶微分光谱建立的模型均不如原始光谱建立的模型效果好。最终建立三个品种枇杷样品的原始光谱在全波段范围经17点平滑后的PLSR模型，模型的校正集和预测集的相关系数分别为0．96和0．95。研究表明近红外光谱检测技术可用于枇杷可溶性固形物含量的定量分析。

The objective of the present research was to evaluate the potential of NIR diffuse reflectance spectroscopy as a way for nondestructive measurement of loquat soluble solid content of three varieties harvested from two orchards. According to the analysis, it was shown that the correlation coefficient of curves was relatively high in the two regions of 1 400-1 500 nm and 1 900-2 000 nm. Statistical models were developed using partial least square regression （PLSR）, stepwise multiple linear regression （SMLR） and principal component regression （PCR） methods in the full spectral region from 800 to 2 500 nm and in the combined region of 1 400-1 500 nm and 1 900-2 000 ran. The results of PLSR model in the full spectral region were better than those of other models. The modeling results based on derivative spectra were not as good as those based on original spectra. Finally, a model was established based on the original spectra with 17 points smoothing for all the three varieties in the full spectral region by PLSR method. The correlation coefficients of calibration and validation were 0. 96 and 0.95, respectively. The results indicate that it is feasible to use NIR spectroscopy technique for quantitative analysis of loquat soluble solid content.