Long-term reproducibility detection method for quantitative LIBS using Kalman filtering

Long-term reproducibility is defined as the dispersion of measurement results over a few days with the same operator, equipment, and samples. It is one of the important obstacles faced by the development of laser-induced breakdown spectroscopy (LIBS) technology. The Kalman filtering algorithm is proposed to improve the long-term reproducibility of LIBS quantitative analysis. Based on 32 lower-alloy steel reference samples, the internal calibration curves of six elements, including Mn, Si, Cr, Ni, Ti, and Al, were established based on the standard single-pulse LIBS equipment. The six test samples were then quantitatively verified every 24 hours for ten days using the above calibration curves. The relative standard deviations (RSDs) of the predicted contents of the six elements before and after Kalman filtering were calculated based on 100 filters, and the improvement effect of Kalman filtering on the long-term reproducibility of quantitative LIBS was evaluated. The results showed that the Kalman filtering method could effectively correct the influence of the instrument drift on the long-term reproducibility of LIBS, and the RSDs of repeated measurements of Mn, Si, Cr, Ni, Ti, and Al elements were reduced from 35%, 49%, 63%, 64%, 66%, and 53% to 11%, 18%, 21%, 25%, 37%, and 27%, respectively. The method proposed in this study offers a new way to improve the long-term reproducibility of LIBS. The Kalman filtering method can effectively correct the influence of instrument drift and improve the long-term reproducibility of LIBS.