Detection of K in soil using time-resolved laser-induced breakdown spectroscopy based on convolutional neural networks

Chengxu LU (吕程序); Bo WANG (王博); Xunpeng JIANG (姜训鹏); Junning ZHANG (张俊宁); Kang NIU (牛康); Yanwei YUAN (苑严伟)

doi:10.1088/2058-6272/aaef6e

Plasma Science and Technology > 2019 > 21(3): 34014-034014. > DOI: 10.1088/2058-6272/aaef6e

Chengxu LU (吕程序), Bo WANG (王博), Xunpeng JIANG (姜训鹏), Junning ZHANG (张俊宁), Kang NIU (牛康), Yanwei YUAN (苑严伟). Detection of K in soil using time-resolved laser-induced breakdown spectroscopy based on convolutional neural networks[J]. Plasma Science and Technology, 2019, 21(3): 34014-034014. DOI: 10.1088/2058-6272/aaef6e

Citation:

PDF (1159 KB)

Detection of K in soil using time-resolved laser-induced breakdown spectroscopy based on convolutional neural networks

¹ Chinese Academy of Agricultural Mechanization Sciences, Beijing 100083, People’s Republic of China
² Jiangsu University, Zhenjiang 212013, People’s Republic of China
³ COFCO Feed Co. Ltd, Beijing 100029, People’s Republic of China

Funds: This work is supported by National Natural Science Foundation of China (Grant No. 61505253) and National Key Research and Development Plan of China (Project No. 2016YFD0200601).

More Information

Received Date: July 30, 2018

Graphical Abstract

Abstract

Abstract

One of the technical bottlenecks of traditional laser-induced breakdown spectroscopy (LIBS) is the difficulty in quantitative detection caused by the matrix effect. To troubleshoot this problem, this paper investigated a combination of time-resolved LIBS and convolutional neural networks (CNNs) to improve K determination in soil. The time-resolved LIBS contained the information of both wavelength and time dimension. The spectra of wavelength dimension showed the characteristic emission lines of elements, and those of time dimension presented the plasma decay trend. The one-dimensional data of LIBS intensity from the emission line at 766.49 nm were extracted and correlated with the K concentration, showing a poor correlation of R²_c=0.0967, which is caused by the matrix effect of heterogeneous soil. For the wavelength dimension, the two-dimensional data of traditional integrated LIBS were extracted and analyzed by an artificial neural network (ANN), showing R²_v=0.6318 and the root mean square error of validation (RMSEV)=0.6234. For the time dimension, the two-dimensional data of time-decay LIBS were extracted and analyzed by ANN, showing R²_v=0.7366 and RMSEV=0.7855. These higher determination coefficients reveal that both the non-K emission lines of wavelength dimension and the spectral decay of time dimension could assist in quantitative detection of K. However, due to limited calibration samples, the two-dimensional models presented over-fitting. The three-dimensional data of time-resolved LIBS were analyzed by CNNs, which extracted and integrated the information of both the wavelength and time dimension, showing the R²_v=0.9968 and RMSEV=0.0785. CNN analysis of time-resolved LIBS is capable of improving the determination of K in soil.

FullText(HTML)

References (23)

References

[1]	Chérel I et al 2014 J. Exp. Bot. 65 833
[2]	ISO 2009 Soil quality—Determination of trace elements in extracts of soil by ICP –AES: ISO 22036 2008 https://iso. org/standard/40653.html
[3]	ISO 2013 Soil quality—Determination of trace elements using ICP-MS: ISO/TS 16965 2013 https://iso.org/standard/ 58056.html
[4]	Hussain T et al 2007 Environ. Monit. Assess. 124 131
[5]	Pareja J et al 2013 Appl. Opt. 52 2470
[6]	Meng D S et al 2014 Chin. J. Lasers 41 0515003 (in Chinese)
[7]	Dong D M et al 2013 Spectrosc. Spectr. Anal. 33 785 (in Chinese)
[8]	Zhang J N et al 2014 Trans. Chin. Soc. Agric. Mach. 45 294 (in Chinese)
[9]	Yu K et al 2017 Spectrosc. Spectr. Anal. 37 2879 (in Chinese)
[10]	Guezenoc J et al 2017 Spectrochim. Acta Part B 134 6
[11]	El Rakwe M et al 2017 J. Chemom. 31 e2869
[12]	Noll R 2012 Laser-Induced Breakdown Spectroscopy (Berlin, Heidelberg: Springer)
[13]	Bredice F et al 2017 Spectrochim. Acta Part B 135 48
[14]	Rawat W and Wang Z H 2017 Neur. Comput. 29 2352
[15]	Yao G L, Lei T and Zhong J D 2018 Patt. Recognit. Lett. (https://doi.org/10.1016/j.patrec.2018.05.018)
[16]	Liu J C et al 2017 Analyst 142 4067
[17]	Hiwa S et al 2016 Comput. Intell. Neurosci. 2016 1841945
[18]	Windrim L et al 2017 IEEE Trans. Geosci. Remote Sens. 56 2798
[19]	Snee R D 1977 Technometrics 19 415
[20]	Wang C et al 2018 Spectrosc. Spectr. Anal. 38 36 (in Chinese)
[21]	Kim P 2017 MATLAB Deep Learning: with Machine Learning, Neural Networks and Artificial Intelligence (Berkeley, CA: Springer)
[22]	Glotot X and Bengio Y 2010 Understanding the difficulty of training deep feedforward neural networks Proc. of the 13th Int. Conf. on Artificial Intelligence and Statistics (Chia Laguna Resort, Sardinia, Italy: AISTATS)
[23]	Maas A L, Hannun A Y and Ng A Y 2013 Rectifier nonlinearities improve neural network acoustic models Proc. of the 30th Int. Conf. on Machine Learning Workshop on Deep Learning (Atlanta, GA, USA: ICML)

[1]	Ming SUN (孙明), Zhan TAO (陶瞻), Zhipeng ZHU (朱志鹏), Dong WANG (王东), Wenjun PAN (潘文军). Spectroscopic diagnosis of plasma in atmospheric pressure negative pulsed gas-liquid discharge with nozzle-cylinder electrode[J]. Plasma Science and Technology, 2018, 20(5): 54005-054005. DOI: 10.1088/2058-6272/aab601
[2]	Xinlei ZHU (朱鑫磊), Liancheng ZHANG (张连成), Yifan HUANG (黄逸凡), Jin WANG (王晋), Zhen LIU (刘振), Keping YAN (闫克平). The effect of the configuration of a single electrode corona discharge on its acoustic characteristics[J]. Plasma Science and Technology, 2017, 19(7): 75403-075403. DOI: 10.1088/2058-6272/aa6716
[3]	LU Yijia (路益嘉), JI Linhong (季林红), CHENG Jia (程嘉). Simulation of Dual-Electrode Capacitively Coupled Plasma Discharges[J]. Plasma Science and Technology, 2016, 18(12): 1175-1180. DOI: 10.1088/1009-0630/18/12/06
[4]	QI Xiaohua (齐晓华), YANG Liang (杨亮), YAN Huijie (闫慧杰), JIN Ying (金英), HUA Yue (滑跃), REN Chunsheng (任春生). Experimental Study on Surface Dielectric Barrier Discharge Plasma Actuator with Different Encapsulated Electrode Widths for Airflow Control at Atmospheric Pressure[J]. Plasma Science and Technology, 2016, 18(10): 1005-1011. DOI: 10.1088/1009-0630/18/10/07
[5]	WANG Yanhui (王艳辉), YE Huanhuan (叶换换), ZHANG Jiao (张佼), WANG Qi (王奇), ZHANG Jie (张杰), WANG Dezhen (王德真). Numerical Study of Pulsed Dielectric Barrier Discharge at Atmospheric Pressure Under the Needle-Plate Electrode Configuration[J]. Plasma Science and Technology, 2016, 18(5): 478-484. DOI: 10.1088/1009-0630/18/5/06
[6]	REN Jingyu (任景俞), WANG Tiecheng (王铁成), QU Guangzhou (屈广周), LIANG Dongli (梁东丽), HU Shibin (呼世斌). Evaluation and Optimization of Electrode Configuration of Multi-Channel Corona Discharge Plasma for Dye-Containing Wastewater Treatment[J]. Plasma Science and Technology, 2015, 17(12): 1053-1060. DOI: 10.1088/1009-0630/17/12/13
[7]	WANG Xiaoping(王小平), LI Zhongjian(李中坚), ZHANG Xingwang(张兴旺), LEI Lecheng(雷乐成). Characteristics of Electrode-Water-Electrode Discharge and its Application to Water Treatment[J]. Plasma Science and Technology, 2014, 16(5): 479-485. DOI: 10.1088/1009-0630/16/5/07
[8]	GONG Jianying (巩建英), ZHANG Xingwang (张兴旺), WANG Xiaoping (王小平), LEI Lecheng (雷乐成). Oxidation of S(IV) in Seawater by Pulsed High Voltage Discharge Plasma with TiO ^₂ /Ti Electrode as Catalyst[J]. Plasma Science and Technology, 2013, 15(12): 1209-1214. DOI: 10.1088/1009-0630/15/12/09
[9]	A. A. AZOOZ, M. A. AHMAD. The Effect of the Earthed Electrode Size on the Ignition Voltage of Low-Pressure RF Capacitive Discharge in Argon[J]. Plasma Science and Technology, 2013, 15(9): 881-884. DOI: 10.1088/1009-0630/15/9/09
[10]	LIU Wenzheng (刘文正), ZHANG Dejin (张德金), KONG Fei (孔飞). The Impact of Electrode Configuration on Characteristics of Vacuum Discharge Plasma[J]. Plasma Science and Technology, 2012, 14(2): 122-128. DOI: 10.1088/1009-0630/14/2/08