Spectroscopy and Spectral Analysis, Volume. 41, Issue 1, 188(2021)
COD Concentration Prediction Model Based on Multi-Spectral Data Fusion and GANs Algorithm
Figures & Tables(6)
Fig. 1. COD water sample spectra
(a): UV spectrum; (b): NIR spectrum
Fig. 2. COD prediction model based on GANs algorithms
(a): Data level fusion; (b): Feature level fusion
Fig. 3. Comparison of evaluation parameters of different quantitative prediction models
(a): Correction sets; (b): Verification sets
Table 1. Data fusion and feature level fusion GANs models without normalized treatment
View tableView in ArticleTable 1. Data fusion and feature level fusion GANs models without normalized treatment
融合方法 预处理 特征提取 筛选波段/nm 校正集 验证集 R 2RMSECV 偏差 R 2RMSEP 偏差 LLDF Der-1S-G - - 0.978 2.356 0.318 0.915 3.928 0.764 MLDF Der-1S-G BiPLS 820~952
1 719~1 8360.984 1.659 0.061 0.933 2.013 0.874
Table 2. Statistics of the result of GAN prediction model using different normalization methods
View tableView in ArticleTable 2. Statistics of the result of GAN prediction model using different normalization methods
特征提
取方法归一化方法 校正集 检验集 R 2RMSECV 偏差 R 2RMSECV 偏差 BiPLS SNV 0.989
0.9891.477
1.4670.014
0.0550.973
0.9731.811
2.0120.156
0.259Min-Max-Nor 0.993
0.9941.136
1.0690.023
0.0310.982
0.9851.655
1.8140.211
0.185VN 0.987 1.653 0.063 0.935 2.928 0.851
Table 3. Evaluation parameters of different quantitative prediction models
View tableView in ArticleTable 3. Evaluation parameters of different quantitative prediction models
最优模型分类 校正集 验证集 加标回收率 R 2RMSECV 偏差 R 2RMSEP 偏差 UV全波段 0.944 1 2.817 0.123 0.894 1 3.749 0.801 90.2~121.6 NIR全波段 0.904 6 3.522 0.147 0.877 4 4.110 0.925 82.1~130.4 UV+NIR(LLDF)融合 0.994 7 1.858 -0.035 0.945 7 2.974 -0.128 93.2~110.4 UV+NIR(MLDF)融合 0.997 7 0.976 0.016 0.994 7 1.325 0.059 98.4~103.1
Tools
Get Citation
Copy Citation Text
Ying CHEN, Yang-mei XU, Yuan-jian DI, Xing-ning CUI, Jie ZHANG, Xin-de ZHOU, Chun-yan XIAO, Shao-hua LI. COD Concentration Prediction Model Based on Multi-Spectral Data Fusion and GANs Algorithm[J]. Spectroscopy and Spectral Analysis, 2021, 41(1): 188
Paper Information
Category: Research Articles
Received: Dec. 27, 2018
Accepted: --
Published Online: Apr. 8, 2021
The Author Email:
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20