Fig. 1. Preprocessing results of the spectrum. (a) Interception of the spectrum; (b) smoothing processing; (c) baseline correction; (d) normalization processing

Fig. 2. Flow chart of modeling and forecasting

Fig. 3. Two similar spectral curves

Fig. 4. Spectra of CV,R6G and NB

Fig. 5. Spectra of multi-component and single-component solutions. (a) Preprocessed spectrum; (b) peak-finding result of maximum value method; (c) secondary smoothing result of spectrum; (d) peak-finding result of characteristic peak discrimination algorithm

Fig. 6. Clustering diagram of different features

Fig. 7. Change curve of the peak intensity

Fig. 8. Processing results of the PCA. (a) Cumulative contribution graph of principal components; (b) scatter plot and confidence ellipsoid of the scores of the first 3 principal components

Fig. 9. Cumulative contribution of principal components of multi-component samples. (a) NB, CV; (b) CV, R6G

Fig. 10. Distribution point diagram of the principal components of multi-component sample. (a) NB、CV; (b) CV、R6G

Fig. 11. Prediction results of the rough classification model

Fig. 12. Prediction results of the subdivision model. (a) CV, R6G; (b) NB, R6G; (c) CV, NB

Fig. 13. Correlation curve between true value and predicted value. (a) R6G; (b) NB; (c) CV