A model combining spectrum standardization and dominant factor based partial least square method for carbon analysis in coal by laser-induced breakdown spectroscopy

摘要

Successful quantitative measurement of carbon content in coal usinglaser-induced breakdown spectroscopy (LIBS) is suffered from relatively lowprecision and accuracy. In the present work, the spectrum standardizationmethod was combined with the dominant factor based partial least square (PLS)method to improve the measurement accuracy of carbon content in coal by LIBS.The combination model employed the spectrum standardization method to convertthe carbon line intensity into standard state for more accurately calculatingthe dominant carbon concentration, and then applied PLS with full spectruminformation to correct the residual errors. The combination model was appliedto the measurement of carbon content for 24 bituminous coal samples. Theresults demonstrated that the combination model could further improve themeasurement accuracy compared with both our previously established spectrumstandardization model and dominant factor based PLS model using spectral areanormalized intensity for the dominant factor model. For example, thecoefficient of determination (R2), the root-mean-square error of prediction(RMSEP), and the average relative error (ARE) for the combination model were0.99, 1.75%, and 2.39%, respectively; while those values for the spectrumstandardization method were 0.83, 2.71%, and 3.40%, respectively; and thosevalues for the dominant factor based PLS model were 0.99, 2.66%, and 3.64%,respectively.