Open AccessAcidity measurement of iron ore powders using laser-induced breakdown spectroscopy with partial least squares regression
Author(s) -
Hao Zeng,
C. M. Li,
Mohan Shen,
Xin Yang,
K. H. Li,
Lianbo Guo,
X. Y. Li,
Yongfeng Lu,
Xiaoyan Zeng
Publication year - 2015
Publication title -
optics express
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.394
H-Index - 271
ISSN - 1094-4087
DOI - 10.1364/oe.23.007795
Subject(s) - partial least squares regression , laser induced breakdown spectroscopy , calibration , analytical chemistry (journal) , mean squared error , matrix (chemical analysis) , spectroscopy , materials science , linear regression , chemistry , mineralogy , mathematics , statistics , chromatography , composite material , physics , quantum mechanics
Laser-induced breakdown spectroscopy (LIBS) with partial least squares regression (PLSR) has been applied to measuring the acidity of iron ore, which can be defined by the concentrations of oxides: CaO, MgO, Al₂O₃, and SiO₂. With the conventional internal standard calibration, it is difficult to establish the calibration curves of CaO, MgO, Al₂O₃, and SiO₂ in iron ore due to the serious matrix effects. PLSR is effective to address this problem due to its excellent performance in compensating the matrix effects. In this work, fifty samples were used to construct the PLSR calibration models for the above-mentioned oxides. These calibration models were validated by the 10-fold cross-validation method with the minimum root-mean-square errors (RMSE). Another ten samples were used as a test set. The acidities were calculated according to the estimated concentrations of CaO, MgO, Al₂O₃, and SiO₂ using the PLSR models. The average relative error (ARE) and RMSE of the acidity achieved 3.65% and 0.0048, respectively, for the test samples.