多年份苹果糖度近红外预测模型建立

Developing of predictive model of soluble solids content of apple harvested at different years

中文关键词:  近红外光谱  信息区间优化  苹果  糖度

英文关键词:Near infrared spectroscopy  Informative spectral region optimization  Apple  Soluble solids content

中文摘要:

      目的 开发多年份苹果糖度预测模型。方法 将移动窗口偏最小二乘法（MWPLS）用于优化3个采收年份的苹果糖度信息区间，构建一种新颖的线性组合权重偏最小二乘法(LCW-PLS)模型。结果MWPLS选择结果为4328~4787 cm-1、5323~5512 cm-1、5982~7135 cm-1和7159~7463 cm-1，当对应权重为0.004、0.070、0.066和0.860时，所建LCW-MWPLS模型预测性能较好，其RP = 0.942、RMSEP = 0.649 %Brix和Q = 0.890。结论 LCW-PLS法可改进常规PLS模型，为果品品质分级提供了一种建模参考方法。

英文摘要:

      Objective To develop a general soluble solids content (SSC) model for apple harvested at different years. Methods Moving window partial least squares (MWPLS) were used to optimize informative spectral regions from FT-NIR spectra. A novel potential method, linear combination weight PLS (LCW-PLS) model, was applied for improving the performance of routine PLS model based on selected informative regions. Results The best calibration model of SSC in apple was obtained by LCW-MWPLS method based on informative spectral regions of 4328?4787, 5323?5512, 5982?7135 and 7159?7463 cm-1 selected by MWPLS procedure, and corresponding weights of 0.004, 0.070, 0.066 and 0.860, respectively. When the performance of LCW-MWPLS model was evaluated by the samples in prediction set, the RP, RMSEP and Q value were 0.942, 0.649 and 0.890, respectively. Conclusion The LCW-MWPLS model giving a prediction error equal to 4% of fresh weight was sufficiently accurate to determine the SSC of apple nondestructively.curate to determine the SSC of apple nondestructively.