Quantitative analysis modeling for the ChemCam spectral data based on laser-induced breakdown spectroscopy using convolutional neural network

Xueqiang CAO (曹学强); Li ZHANG (张立); Zhongchen WU (武中臣); Zongcheng LING (凌宗成); Jialun LI (李加伦); Kaichen GUO (郭恺琛)

doi:10.1088/2058-6272/aba5f6

Plasma Science and Technology > 2020 > 22(11): 115502. > DOI: 10.1088/2058-6272/aba5f6

Xueqiang CAO (曹学强), Li ZHANG (张立), Zhongchen WU (武中臣), Zongcheng LING (凌宗成), Jialun LI (李加伦), Kaichen GUO (郭恺琛). Quantitative analysis modeling for the ChemCam spectral data based on laser-induced breakdown spectroscopy using convolutional neural network[J]. Plasma Science and Technology, 2020, 22(11): 115502. DOI: 10.1088/2058-6272/aba5f6

Citation:

PDF (1495 KB)

Quantitative analysis modeling for the ChemCam spectral data based on laser-induced breakdown spectroscopy using convolutional neural network

¹ School of Mechanical, Electrical & Information Engineering, Shandong University, Weihai 264209, People's Republic of China
² Shandong Provincial Key Laboratory of Optical Astronomy and Solar-Terrestrial Environment, Institute of Space Sciences, Shandong University, Weihai 264209, People's Republic of China

Funds: This work is supported by the Pre-research project on Civil Aerospace Technologies (No. D020102) funded by China National Space Administration (CNSA). We also thank the funding from National Natural Science Foundation of China (Nos. U1931211 and 41573056), the Natural Science Foundation of Shandong Province (No. ZR2019MD008) and the Major Research Project of Shandong Province (No. GG201809130208).

More Information

Received Date: March 26, 2020
Revised Date: July 12, 2020
Accepted Date: July 13, 2020

Graphical Abstract

Abstract

Abstract

Laser-induced breakdown spectroscopy (LIBS) has been applied to many fields for the quantitative analysis of diverse materials. Improving the prediction accuracy of LIBS regression models is still of great significance for the Mars exploration in the near future. In this study, we explored the quantitative analysis of LIBS for the one-dimensional ChemCam (an instrument containing a LIBS spectrometer and a Remote Micro-Imager) spectral data whose spectra are produced by the ChemCam team using LIBS under the Mars-like atmospheric conditions. We constructed a convolutional neural network (CNN) regression model with unified parameters for all oxides, which is efficient and concise. CNN that has the excellent capability of feature extraction can effectively overcome the chemical matrix effects that impede the prediction accuracy of regression models. Firstly, we explored the effects of four activation functions on the performance of the CNN model. The results show that the CNN model with the hyperbolic tangent (tanh) function outperforms the CNN models with the other activation functions (the rectified linear unit function, the linear function and the Sigmoid function). Secondly, we compared the performance among the CNN models using different optimization methods. The CNN model with the stochastic gradient descent optimization and the initial learning rate=0.0005 achieves satisfactory performance compared to the other CNN models. Finally, we compared the performance of the CNN model, the model based on support vector regression (SVR) and the model based on partial least square regression (PLSR). The results exhibit the CNN model is superior to the SVR model and the PLSR model for all oxides. Based on the above analysis, we conclude the CNN regression model can effectively improve the prediction accuracy of LIBS.
- laser-induced breakdown spectroscopy,
- convolutional neural network,
- activationfunction,
- optimization method,
- quantitative analysis

FullText(HTML)

References (32)

References

[1]	Rao A P et al 2019 Atoms 7 84
[2]	Cremers D A et al 2012 Appl. Spectrosc. 66 250
[3]	Shattan M B et al 2017 Appl. Opt. 56 9868
[4]	Thomas N H et al 2018 J. Geophys. Res.-Planets 123 1996
[5]	Maurice S et al 2016 J. Anal. Atom. Spectrom. 31 863
[6]	Grotzinger J P et al 2012 J. Space Sci. Rev. 170 5
[7]	Le Mouélic S et al 2015 Icarus 249 93
[8]	Murdoch N et al 2019 Planet. Space Sci. 165 260
[9]	Cai T N et al 2019 Spectrosc. Spect. Anal. 39 1623 (in Chinese)
[10]	Zhang B H et al 2019 Sensors 19 1727
[11]	Sanghapi H K et al 2016 Spectrochim. Acta B 122 9
[12]	Barimani S and Kleinebudde P 2017 Eur. J. Pharm. Biopharm.119 28
[13]	Liu F et al 2018 Sensors 18 705
[14]	Anderson R B et al 2017 Spectrochim. Acta B 129 49
[15]	Duan H W, Han L J and Huang G Q 2019 Molecules 24 3753
[16]	Wang T et al 2018 Spectrochim. Acta B 149 300
[17]	Shi Q et al 2015 J. Anal. Atom. Spectrom. 30 2384
[18]	Dyar M D et al 2012 Spectrochim. Acta B 70 51
[19]	Zhang X L et al 2019 Anal. Chim. Acta 1058 48
[20]	Clegg S M et al 2017 Spectrochim. Acta B 129 64
[21]	Wiens R C et al 2013 Spectrochim. Acta B 82 1
[22]	Krizhevsky A, Sutskever I and Hinton G E 2017 Commun.ACM 60 84
[23]	Muhammad K et al 2018 IEEE Access 6 18174
[24]	Manit J et al 2020 IET Biometrics 9 31
[25]	Lu C X et al 2019 Plasma Sci. Technol. 21 34014
[26]	Ioffe S and Szegedy C 2015 Batch normalization: accelerating deep network training by reducing internal covariate shift Proc. 32nd Int. Conf. on Machine Learning (Lille, France) (IMLS) p 448
[27]	Glorot X and Bengio Y 2010 Understanding the difficulty of training deep feedforward neural networks Proc. 13th Int.Conf. on Artificial Intelligence and Statistics (Chia Laguna Resort, Sardinia Italy) (AISTATS) p 249
[28]	He K M et al 2015 Delving deep into rectifiers: surpassing human-level performance on ImageNet classification Proc.2015 IEEE Int. Conf. on Computer Vision (Santiago, Chile:IEEE) p 1026
[29]	Darken C et al 1992 Learning rate schedules for faster stochastic gradient search Neural Networks for Signal Processing II Proc. 1992 IEEE Workshop (Helsingoer,Denmark) (IEEE) p 3
[30]	Kingma D P and Ba J L 2015 Adam: a method for stochastic optimization Proc. 3rd Int. Conf. on Learning Representations (San Diego, USA) (ICLR) p 1
[31]	Vapnik V N 1999 IEEE Trans. Neural Netw. Learn. 10 988
[32]	Wold S, Sjöström M and Eriksson L 2001 Chemometr. Intell.Lab. 58 109

[1]	Weiwei HAN, Duixiong SUN, Guoding ZHANG, Honglin WANG, Kai GUO, Yuzhuo ZHANG, Haoliang WANG, Denghong ZHANG, Chenzhong DONG, Maogen SU. Research on batch multielement rapid quantitative analysis based on the standard curve-assisted calibration-free laser-induced breakdown spectroscopy method[J]. Plasma Science and Technology, 2024, 26(9): 095502. DOI: 10.1088/2058-6272/ad5119
[2]	Guodong WANG (王国栋), Lanxiang SUN (孙兰香), Wei WANG (汪为), Tong CHEN (陈彤), Meiting GUO (郭美亭), Peng ZHANG (张鹏). A feature selection method combined with ridge regression and recursive feature elimination in quantitative analysis of laser induced breakdown spectroscopy[J]. Plasma Science and Technology, 2020, 22(7): 74002-074002. DOI: 10.1088/2058-6272/ab76b4
[3]	Yaguang MEI (梅亚光), Shusen CHENG (程树森), Zhongqi HAO (郝中骐), Lianbo GUO (郭连波), Xiangyou LI (李祥友), Xiaoyan ZENG (曾晓雁), Junliang GE (葛军亮). Quantitative analysis of steel and iron by laser-induced breakdown spectroscopy using GA-KELM[J]. Plasma Science and Technology, 2019, 21(3): 34020-034020. DOI: 10.1088/2058-6272/aaf6f3
[4]	Jiamin LIU (刘佳敏), Ding WU (吴鼎), Cailong FU (付彩龙), Ran HAI (海然), Xiao YU (于潇), Liying SUN (孙立影), Hongbin DING (丁洪斌). Improvement of quantitative analysis of molybdenum element using PLS-based approaches for laser-induced breakdown spectroscopy in various pressure environments[J]. Plasma Science and Technology, 2019, 21(3): 34017-034017. DOI: 10.1088/2058-6272/aaf821
[5]	Xiaomeng LI (李晓萌), Huili LU (陆慧丽), Jianhong YANG (阳建宏), Fu CHANG (常福). Semi-supervised LIBS quantitative analysis method based on co-training regression model with selection of effective unlabeled samples[J]. Plasma Science and Technology, 2019, 21(3): 34015-034015. DOI: 10.1088/2058-6272/aaee14
[6]	Congyuan PAN (潘从元), Jiao HE (何娇), Guangqian WANG (王广谦), Xuewei DU (杜学维), Yongbin LIU (刘永斌), Yahui SU (苏亚辉). An efficient procedure in quantitative analysis using laser-induced breakdown spectroscopy[J]. Plasma Science and Technology, 2019, 21(3): 34012-034012. DOI: 10.1088/2058-6272/aaf50f
[7]	Minchao CUI (崔敏超), Yoshihiro DEGUCHI (出口祥啓), Zhenzhen WANG (王珍珍), Seiya TANAKA (田中诚也), Min-Gyu JEON (全敏奎), Yuki FUJITA (藤田裕贵), Shengdun ZHAO (赵升吨). Remote open-path laser-induced breakdown spectroscopy for the analysis of manganese in steel samples at high temperature[J]. Plasma Science and Technology, 2019, 21(3): 34007-034007. DOI: 10.1088/2058-6272/aaeba7
[8]	Qingdong ZENG (曾庆栋), Fan DENG (邓凡), Zhiheng ZHU (朱志恒), Yun TANG (唐云), Boyun WANG (王波云), Yongjun XIAO (肖永军), Liangbin XIONG (熊良斌), Huaqing YU (余华清), Lianbo GUO (郭连波), Xiangyou LI (李祥友). Portable fiber-optic laser-induced breakdown spectroscopy system for the quantitative analysis of minor elements in steel[J]. Plasma Science and Technology, 2019, 21(3): 34006-034006. DOI: 10.1088/2058-6272/aadede
[9]	Dan LUO (罗丹), Ying LIU (刘英), Xiangyu LI (李香宇), Zhenyang ZHAO (赵珍阳), Shigong WANG (王世功), Yong ZHANG (张勇). Quantitative analysis of C, Si, Mn, Ni, Cr and Cu in low-alloy steel under ambient conditions via laser-induced breakdown spectroscopy[J]. Plasma Science and Technology, 2018, 20(7): 75504-075504. DOI: 10.1088/2058-6272/aabc5d
[10]	WANG Shaolong (王绍龙), WANG Yangen (王阳恩), CHEN Shanjun (陈善俊), CHEN Qi (陈奇). Quantitative Analysis of Mg in Pipeline Dirt Based on Laser-Induced Breakdown Spectroscopy[J]. Plasma Science and Technology, 2015, 17(8): 716-720. DOI: 10.1088/1009-0630/17/8/18

Cited By

Cited by

Periodical cited type(21)

1.	Pan, J., Qiao, X., Zhang, C. et al. Stacking Ensemble Learning-Assisted Simulation of Plasma-Catalyzed CO2 Reforming of Methane. Electronics (Switzerland), 2025, 14(7): 1329. DOI:10.3390/electronics14071329
2.	Liu, S., Ren, Z., Wang, W. et al. The instability prediction of non-resonant energetic particle modes based on machine learning algorithms. Physica Scripta, 2025, 100(3): 036007. DOI:10.1088/1402-4896/adafe4
3.	Yan, X.-T., Bao, N.-N., Zhao, C.-Y. et al. NTVTOK-ML: Fast surrogate model for neoclassical toroidal viscosity torque calculation in tokamaks based on machine learning methods. Computer Physics Communications, 2025. DOI:10.1016/j.cpc.2024.109413
4.	Li, H., Wang, L., Fu, Y.L. et al. Surrogate model of turbulent transport in fusion plasmas using machine learning. Nuclear Fusion, 2025, 65(1): 016015. DOI:10.1088/1741-4326/ad8b5b
5.	Liu, Z.Y., Qiu, H.R., Fu, G.Y. et al. Prediction of fishbone linear instability in tokamaks with machine learning methods. Nuclear Fusion, 2025, 65(1): 016007. DOI:10.1088/1741-4326/ad8d69
6.	Xu, J., Luan, Q., Li, H. et al. Neural network based fast prediction of double tearing modes in advanced tokamak plasmas. Physics of Plasmas, 2024, 31(12): 122113. DOI:10.1063/5.0229910
7.	Zheng, G.H., Yang, Z.Y., Liu, S.F. et al. Real-time equilibrium reconstruction by multi-task learning neural network based on HL-3 tokamak. Nuclear Fusion, 2024, 64(12): 126041. DOI:10.1088/1741-4326/ad8014
8.	Zhang, H., Zhou, L., Liu, Y. et al. Deep learning approaches to recover the plasma current density profile from the safety factor based on Grad-Shafranov solutions across multiple tokamaks. Plasma Science and Technology, 2024, 26(5): 055101. DOI:10.1088/2058-6272/ad13e3
9.	Li, G., Li, L., Qiao, X. et al. Research on Machine Learning-Assisted Pulse Discharge Plasma Catalysis for Methane-Carbon Dioxide Simulation. 2024. DOI:10.1109/ICCSNT62291.2024.10776678
10.	Liu, Y., Li, J. Gyro-Landau-fluid simulations of impurity effects on ion temperature gradient driven turbulence transport. Plasma Science and Technology, 2024, 26(1): 015101. DOI:10.1088/2058-6272/ad0c9b
11.	Li, H., Fu, Y.-L., Li, J.-Q. et al. Simulation Prediction of Heat Transport with Machine Learning in Tokamak Plasmas. Chinese Physics Letters, 2023, 40(12): 125201. DOI:10.1088/0256-307X/40/12/125201
12.	Li, H., Li, J.-Q., Wang, Z.-X. Global Effects on Drift Wave Microturbulence in Tokamak Plasmas. Chinese Physics Letters, 2023, 40(10): 105201. DOI:10.1088/0256-307X/40/10/105201
13.	Wang, Z., Qiu, Z., Wang, L. et al. Summary of the 10th Conference on Magnetically Confined Fusion Theory and Simulation (CMCFTS). Plasma Science and Technology, 2023, 25(8): 081001. DOI:10.1088/2058-6272/acc14d
14.	Liang, C., Huang, D., Lu, S. et al. Determining global property of dusty plasma from single particle dynamics using machine learning. Physical Review Research, 2023, 5(3): 033086. DOI:10.1103/PhysRevResearch.5.033086
15.	Pan, J., Liu, Y., Zhang, S. et al. Deep learning-assisted pulsed discharge plasma catalysis modeling. Energy Conversion and Management, 2023. DOI:10.1016/j.enconman.2022.116620
16.	Zheng, W., Xue, F., Shen, C. et al. Overview of machine learning applications in fusion plasma experiments on J-TEXT tokamak. Plasma Science and Technology, 2022, 24(12): 124003. DOI:10.1088/2058-6272/ac9e46
17.	Wai, J.T., Boyer, M.D., Kolemen, E. Neural net modeling of equilibria in NSTX-U. Nuclear Fusion, 2022, 62(8): 086042. DOI:10.1088/1741-4326/ac77e6
18.	Wang, T., Li, B., Gao, J. et al. Monitoring of two-dimensional tungsten concentration profiles on the HL-2A tokamak. Plasma Physics and Controlled Fusion, 2022, 64(8): 084003. DOI:10.1088/1361-6587/ac77b9
19.	Yang, Y.R., Fu, S.H., Ding, Z.F. Effects of ion extraction on discharges in gridded ion source. AIP Advances, 2022, 12(5): 055325. DOI:10.1063/5.0082813
20.	Ida, K., McDermott, R.M., Holland, C. et al. Joint meeting of 9th Asia Pacific-Transport Working Group (APTWG) & EU-US Transport Task Force (TTF) workshop. Nuclear Fusion, 2022, 62(3): 037001. DOI:10.1088/1741-4326/ac3f19
21.	Li, H., Li, J.Q., Fu, Y.L. et al. Simulation prediction of micro-instability transition and associated particle transport in tokamak plasmas. Nuclear Fusion, 2022, 62(3): 036014. DOI:10.1088/1741-4326/ac486b

Other cited types(0)

Get Citation

PDF

XML

Article views (180) PDF downloads (368) Cited by(21)

Turn off MathJax

Article Contents

Abstract

References

Quantitative analysis modeling for the ChemCam spectral data based on laser-induced breakdown spectroscopy using convolutional neural network