研究目的
To improve the classification accuracy of stainless steel alloys using laser-induced breakdown spectroscopy (LIBS) under highly fluctuating signal conditions by employing an intensity-ratio-based analysis method.
研究成果
The proposed intensity-ratio-based LIBS classification method significantly improves the classification accuracy of stainless steel alloys under fluctuating signal conditions, compared to full-spectra PCA or intensity-based analysis. The method is suitable for industrial scrap sorting systems requiring minimal maintenance and low system price.
研究不足
The study acknowledges the significant fluctuation of LIBS signal intensity due to varying focusing conditions, which could affect classification accuracy. The method's performance under industrial-scale scrap sorting conditions, with scraps of varying shapes and sizes, was not fully explored.
1：Experimental Design and Method Selection:
The study utilized LIBS for the classification of stainless steel alloys, focusing on the selection of spectral line pairs for intensity ratio calculation based on elemental concentration and upper levels of emission lines.
2：Sample Selection and Data Sources:
Certified reference materials (CRMs) of nine different kinds of austenitic stainless steel were used, purchased from the National Institute of Standards and Technology (NIST) and Brammer Standard Company Incorporation.
3：List of Experimental Equipment and Materials:
A flashlamp pumped Q-switched Nd:YAG laser (Nano L 90?100, Litron Lasers) was used for ablation, and a dual channel spectrometer (Avaspec-ULS2048-2-USB2, Avantes) was employed for plasma emission detection.
4：Experimental Procedures and Operational Workflow:
Each CRM sample was analyzed at 8 different positions, with LIBS spectra measured at five different spots per position. The first shot at each spot was excluded to avoid surface contamination effects.
5：Data Analysis Methods:
The classification accuracy was evaluated by comparing the results based on the full-spectra principal component analysis (PCA) and on the intensities of selected lines, using ten-fold cross-validation by linear discriminant analysis.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
