Quantitative Analysis of Liquid Steel Element in LIBS Using SVR Improved by Particle Swarm Optimization

Youliang Yang; Lu Wang; Cuihong Ma

doi:10.3788/LOP57.053002

Journals >Laser & Optoelectronics Progress >Volume 57 >Issue 5 >Page 053002 > Article

Laser & Optoelectronics Progress
Vol. 57, Issue 5, 053002 (2020)

Quantitative Analysis of Liquid Steel Element in LIBS Using SVR Improved by Particle Swarm Optimization

Youliang Yang, Lu Wang^*, and Cuihong Ma

Author Affiliations

College of Electrical Engineering, North China University of Science and Technology, Tangshan, Hebei 063210, China

show less

DOI: 10.3788/LOP57.053002 Cite this Article Set citation alerts

Youliang Yang, Lu Wang, Cuihong Ma. Quantitative Analysis of Liquid Steel Element in LIBS Using SVR Improved by Particle Swarm Optimization[J]. Laser & Optoelectronics Progress, 2020, 57(5): 053002 Copy Citation Text

show less

Fig. 1. LIBS experimental platform

Download full size

Fig. 2. Liquid steel sample

Download full size

Fig. 3. LIBS spectrum of ERM284-2D

Download full size

Fig. 4. Flow chart of SVR and Cat-fish PSO-SVR algorithms

Download full size

Fig. 5. Fitness changing curve of Cat-fish PSO

Download full size

Fig. 6. Results of elemental prediction analysis in molten steel

Download full size

Sample	Fe	Mn	Ni	Cr	Si	Cu	Co	P	S
ERM281-1	70.329	9.370	0.258	0.176	1.929	0.048	0.023	0.012	0.016
ERM282-1	67.150	10.860	0.475	1.109	0.734	0.086	0	0.019	0.004
ERM279-2	75.786	10.603	0.786	2.107	1.516	0.088	0	0	0
ERM284-2D	67.496	10.720	1.744	0.183	2.537	0.020	0.053	0.026	0.024
ERM287-1D	67.439	10.350	1.480	3.203	1.569	0.016	0.148	0.027	0.001
ERM289-1D	55.238	24.680	1.106	0.000	4.531	0.049	0.065	0.012	0.003
ERM292-1D	68.927	10.090	0.976	2.139	2.402	0.037	0.026	0.018	0.006
ERM295-1D	48.360	24.400	0.397	1.481	4.418	0.017	0.045	0.017	0.001
ERM296-1D	82.020	5.790	0.545	0.149	1.242	0.117	0.021	0.018	0.003
ERM298-1D	63.380	7.056	0.316	3.201	4.262	0.016	0.055	0.020	0.001

Table 1. Element composition of steel samples%

Component number	1	2	3	4	5	6	7	8	9
Eigen value	9.674	7.861	5.349	3.245	0.934	0.761	0.670	0.511	0.239
Cumulative variance contribution /%	33.08	26.88	18.29	11.10	3.19	2.60	2.30	1.75	0.82

Table 2. Cumulative variance contribution of each influencing factor

Corrosion factor	PC 1	PC 2	PC 3	PC 4
Peak intensity	0.967	0.377	0.654	0.385
Integrated intensity	0.019	0.392	-0.159	0.959
Signal-to-back ratio	0.069	0.818	0.041	-0.172
Internal standard corrected element concentration	0.507	0.068	0.882	-0.567
Pulse trigger delay /μs	0.297	0.014	0.148	0.069
Distance /mm	0.503	0.552	0.347	0.462
Spectral peak value	-0.435	0.462	0.297	0.293
Spectral efficiency /%	0.514	-0.408	-0.502	0.377
Intensity of temperature thermal radiation /(J·cm^-2·min^-1)	0.205	-0.538	0.527	-0.151

Table 3. Eigen vectors of principal components

Element	SVR model	Cat-fish PSO-SVR model
Element	R²		R_SD /%	R_MSE /%	R²	R_SD /%	R_MSE /%
Mn	0.9102	6.10	3.34	0.9511	2.46	1.07
Ni	0. 9177	6.83	3.18	0.9627	3.17	0.97
Cr	0. 9174	9.52	4.07	0.9845	3.99	1.26
Si	0.9099	8.27	3.02	0.9522	4.50	1.43

Table 4. Regression prediction results for test set

Youliang Yang, Lu Wang, Cuihong Ma. Quantitative Analysis of Liquid Steel Element in LIBS Using SVR Improved by Particle Swarm Optimization[J]. Laser & Optoelectronics Progress, 2020, 57(5): 053002

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information