Quantitative Analysis of Steel Samples Using Laser-Induced Breakdown Spectroscopy with an Artificial Neural Network Incorporating a Genetic Algorithm.

Kuohu Li,Lianbo Guo,Jiaming Li,Xinyan Yang,Rongxing Yi,Xiangyou Li,Yongfeng Lu,Xiaoyan Zeng

DOI: https://doi.org/10.1364/ao.56.000935

IF: 1.9

2017-01-01

Applied Optics

Abstract:In this work, a genetic algorithm (GA) was employed to select the intensity ratios of the spectral lines belonging to the target and domain matrix elements, then these selected line-intensity ratios were taken as inputs to construct an analysis model based on an artificial neural network (ANN) to analyze the elements copper (Cu) and vanadium (V) in steel samples. The results revealed that the root mean square errors of prediction (RMSEPs) for the elements Cu and V can reach 0.0040 wt. % and 0.0039 wt. %, respectively. Compared to 0.0190 wt. % and 0.0201 wt. % of the conventional internal calibration approach, the reduction rates of the RMSEP values reached 78.9% and 80.6%, respectively. These results indicate that the GA combining ANN can excellently execute the quantitative analysis in laser-induced breakdown spectroscopy for steel samples and further improve analytical accuracy.