Quantitative Analysis of Total Phenol Content in Cabernet Sauvignon Grape Based on Near-Infrared Spectroscopy

Fig. 4. Characteristic wavebands selected by CARS algorithm
(a): Variation of the number of selected wavelength variables; (b): Variation of RMSECV;(c): The changing trend of variable regression coefficients

Download full size

View in Article

Fig. 5. (a) Frequency distribution of bands selected by GA; (b) Selection of wavelengths by the GA method

Download full size

View in Article

Fig. 6. Characteristic wavebands selected by si-PLS method

Download full size

View in Article

Fig. 7. Plot of 50 wavelengths selected by SPA

Download full size

View in Article

Fig. 8. The influence of the number of neurons in different hidden layers on the performance of the model

Download full size

View in Article

Fig. 9. Scatter plot of real value and predicted value of total phenol content

Download full size

View in Article

Table 1. Statistical results of total phenol content of Cabernet Sauvignon in different harvest stages (mg·g^-1)
View table
View in Article
Table 1. Statistical results of total phenol content of Cabernet Sauvignon in different harvest stages (mg·g^-1)
WAV 样本数量最大值最小值平均值标准偏差
1 40 28.01 14.65 18.99 2.956
2 40 28.96 17.07 22.20 2.919
3 40 29.64 11.23 24.61 3.366
4 40 34.85 19.86 28.10 3.576
5 40 32.64 17.12 25.68 4.397

Table 2. Results of ELM modeling for the total phenol content with different pre-treatment methods

View table

View in Article

Table 2. Results of ELM modeling for the total phenol content with different pre-treatment methods

预处理方法	校正集		预测集		RPD
预处理方法	R_c	RMSEC	R_p	RMSEP	RPD
原始	0.645 4	3.615 4	0.595 9	3.487 5	1.245 3
MSC	0.688 0	3.494 1	0.680 9	2.943 1	1.365 5
SNV	0.725 4	3.313 9	0.654 3	3.039 1	1.322 3
MC	0.674 5	3.494 7	0.610 1	3.441 0	1.262 1
MA	0.646 4	3.591 3	0.606 9	3.506 8	1.258 2
一阶导数+SG	0.634 2	3.649 2	0.623 1	3.501 3	1.278 6

Table 3. Performance comparison of different activation functions
View table
View in Article
Table 3. Performance comparison of different activation functions
激活函数
类型 RMSEC R_c RMSEP R_p RPD 运行时
间/s
Sigmoid 2.196 0 0.893 3 1.861 5 0.878 3 2.091 4 34.39
Sine 2.602 7 0.846 4 1.946 6 0.866 0 1.999 9 33.84
Hardlim 3.432 4 0.710 8 3.632 4 0.374 4 1.078 5 33.46

Table 4. The prediction results of ELM and GA-ELM models

View table

View in Article

Table 4. The prediction results of ELM and GA-ELM models

Prediction models	Selection methods	Variables number	Calibration sets		Prediction sets		RPD
Prediction models	Selection methods	Variables number	R_c	RMSEC	R_p	RMSEP	RPD
ELM	CARS	69	0.856 3	2.524 2	0.824 5	2.203 2	1.767 0
	GA	168	0.763 8	3.158 8	0.732 8	2.603 8	1.469 5
	Si-PLS	413	0.733 4	3.346 4	0.669 9	2.778 2	1.346 9
	SPA	50	0.742 7	3.216 4	0.731 4	2.847 1	1.466 4
GA-ELM	CARS	69	0.901 7	2.112 4	0.901 3	1.686 8	2.308 0
	GA	168	0.835 6	2.688 8	0.834 7	2.107 3	1.815 8
	Si-PLS	413	0.798 7	2.962 2	0.719 1	2.600 3	1.375 4
	SPA	50	0.854 7	2.493 7	0.844 9	2.233 4	1.869 3

Tools

Get Citation

Copy Citation Text

Yi-jia LUO, He ZHU, Xiao-han LI, Juan DONG, Hao TIAN, Xue-wei SHI, Wen-xia WANG, Jing-tao SUN. Quantitative Analysis of Total Phenol Content in Cabernet Sauvignon Grape Based on Near-Infrared Spectroscopy[J]. Spectroscopy and Spectral Analysis, 2021, 41(7): 2036

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites

Paper Information

Category: Research Articles

Received: Jul. 2, 2020

Accepted: --

Published Online: Sep. 8, 2021

The Author Email: LUO Yi-jia (luoyijiayj@126.com)

DOI:10.3964/j.issn.1000-0593(2021)07-2036-07

Topics

laser devices and laser physics

Lasers and Laser Optics

Laser physics

laser manufacturing

Instrumentation, Measurement and Metrology

Table 1. Statistical results of total phenol content of Cabernet Sauvignon in different harvest stages (mg·g-1)

Table 1. Statistical results of total phenol content of Cabernet Sauvignon in different harvest stages (mg·g-1)

Table 2. Results of ELM modeling for the total phenol content with different pre-treatment methods

Table 2. Results of ELM modeling for the total phenol content with different pre-treatment methods

Table 3. Performance comparison of different activation functions

Table 3. Performance comparison of different activation functions

Table 4. The prediction results of ELM and GA-ELM models

Table 4. The prediction results of ELM and GA-ELM models

Table 1. Statistical results of total phenol content of Cabernet Sauvignon in different harvest stages (mg·g^-1)

Table 1. Statistical results of total phenol content of Cabernet Sauvignon in different harvest stages (mg·g^-1)