Identification and Restoration of Pseudo-Hydrolyzed Animal Protein of Lacteus Camelus Based on iPLS Model of Near-Infrared Measurement Spectrum of 6 mm Detection Plate

Ke-yan YUAN; Rong WANG; Xiang-xiang WANG; Li-ping XUE; Li YU

doi:10.3964/j.issn.1000-0593(2022)10-3143-05

Journals >Spectroscopy and Spectral Analysis >Volume 42 >Issue 10 >Page 3143 > Article

Spectroscopy and Spectral Analysis
Vol. 42, Issue 10, 3143 (2022)

Identification and Restoration of Pseudo-Hydrolyzed Animal Protein of Lacteus Camelus Based on iPLS Model of Near-Infrared Measurement Spectrum of 6 mm Detection Plate

Ke-yan YUAN^1、*, Rong WANG^2、2;, Xiang-xiang WANG^2、2;, Li-ping XUE^2、2;, and Li YU^{2、2; *;}

Author Affiliations

¹1. Huhhot City Inspection and Testing Center, Huhhot 010018, China

²2. School of Environment and Energy Engineering, Anhui Jianzhu University, Hefei 230601, China

show less

DOI: 10.3964/j.issn.1000-0593(2022)10-3143-05 Cite this Article

Ke-yan YUAN, Rong WANG, Xiang-xiang WANG, Li-ping XUE, Li YU. Identification and Restoration of Pseudo-Hydrolyzed Animal Protein of Lacteus Camelus Based on iPLS Model of Near-Infrared Measurement Spectrum of 6 mm Detection Plate[J]. Spectroscopy and Spectral Analysis, 2022, 42(10): 3143 Copy Citation Text

show less

Fig. 1. Calculate RMSECV value of principal component

Download full size | View in the Article

Fig. 2. Global 1 principal component PLS regression curve

Download full size | View in the Article

Fig. 3. Interval division

Download full size | View in the Article

Fig. 4. Interval selection when interval is divided into 30

Download full size | View in the Article

Fig. 5. Regression of the 12th interval model

Download full size | View in the Article

实验号	水解动物蛋白/ [g·(100 mL)^-1]	实验号	水解动物蛋白/ [g·(100 mL)^-1]	实验号	水解动物蛋白/ [g·(100 mL)^-1]	实验号	水解动物蛋白/ [g·(100 mL)^-1]
1	0.01	12	0.50	23	1.05	34	1.60
2	0.03	13	0.55	24	1.10	35	1.65
3	0.05	14	0.60	25	1.15	36	1.70
4	0.10	15	0.65	26	1.20	37	1.75
5	0.15	16	0.70	27	1.25	38	1.80
6	0.20	17	0.75	28	1.30	39	1.85
7	0.25	18	0.80	29	1.35	40	1.90
8	0.30	19	0.85	30	1.40	41	1.95
9	0.35	20	0.90	31	1.45	42	2.00
10	0.40	21	0.95	32	1.50	43	0.00
11	0.45	22	1.00	33	1.55

Table 1. The configuration of experimental samples

View in the Article

样品集	样品数	最低值/[g·(100 mL)^-1]	最高值/[g·(100 mL)^-1]	平均值/[g·(100 mL)^-1]
训练集	33	0.00	2.00	0.95
测试集	10	0.10	1.90	0.97

Table 2. Training set test set condition

View in the Article

光谱预处理方法	主成分数	r	RMSECV
一阶倒数	1	0.119 9	0.670 3
SG平滑	1	0.849 1	0.326 0
SNV	1	0.407 7	0.429 0
一阶导数+SG	1	0.234 4	0.708 6
一阶导数+SNV	1	0.163 4	0.466 7
SG+SNV	1	0.399 9	0.429 8
原始光谱	1	0.854 4	0.322 0

Table 3. Preprocessing method global modeling parameters

View in the Article

区间数量	小于全局RMSECV区间个数	主成分数	子区间号	对应波数	相关系数	区间RMSECV值
15	1	14	6	7 988.16~7 590.87	0.913 1	0.250 5
20	1	7	8	7 887.87~7 590.87	0.941 3	0.208 1
25	1	6	10	7 882.30~7 587.01	0.929 0	0.227 1
30	1	6	12	7 787.56~7 590.87	0.945 1	0.200 1
35	1	6	14	7 737.44~7 567.73	0.939 3	0.211 7
40	1	5	16	7 737.44~7 590.87	0.928 5	0.228 8

Table 4. Table of interval partitioning parameters

Download Citation

Save the article for my favorites

Paper Information