Research And System Integration Of Rare Earth Ores Composition Analysis Method Based On LIBS

Rare earth ores refer to an ore containing a certain amount of rare earth elements.After processing and separation,rare earth elements of various purities can be obtained,which are usually called"rare earth".Rare earth has high application value and plays an important role in modern chemical industry,electronics,steel metallurgy,aerospace,communication,new energy,medicine,biotechnology and other fields.However,the exploitation,processing,and utilization of rare earth ore resources have high technological content and environmental risks.Moreover,the global rare earth ore market supply-demand pattern also faces enormous challenges.Therefore,it is of great significance for the development of rare earth ore resources to explore a method that can quickly detect and analyze elements in rare earth ores on site.In this paper,based on the actual demand of rare earth ore exploration,taking rare earth elements Lu and Y in rare earth ore as the research objects,the quantitative analysis method of rare earth elements in rare earth ore based on laser-induced breakdown spectroscopy(LIBS)combined with chemometrics and system integration was studied.The main research contents are as follows:(1)Taking the rare earth elements Lu and Y in rare earth ores as the research objects,the influence of different pretreatment methods and variable selection methods(i PLS,VIP,i PLS-VIP)on the prediction performance of the PLS calibration model was explored.The results show that the WT-i PLS-VIP-PLS calibration model had the best prediction ability,and the R~2p,RMSEP and MREP of the Lu element prediction set are 0.9897,0.8150μg/g and 7.54%respectively.R~2p,RMSEP and MREP of the Y element prediction set were 0.9833,97.1047μg/g and 7.66%respectively.Compared with the PLS calibration model established by the original spectrum,the mean relative errors(MRE)of rare earth elements Lu and Y were reduced by53.28%and 49.54%respectively.Therefore,LIBS technology combined with WT-i PLS-VIP-PLS can realize the quantitative analysis of rare earth elements Lu and Y in rare earth ores.(2)In order to further improve the calculation efficiency and prediction accuracy of the calibration model,the LIBS spectrum of rare earth ore after wavelet transform(WT)processing was selected by the Grey Wolf Optimization(GWO)variable selection strategy,and the parameters of GWO were optimized by 5-fold cross-validation.Under the optimal parameters of GWO,a quantitative analysis method of rare earth elements Lu and Y based on LIBS technology combined with WT-GWO-PLS was established.The results show that the prediction accuracy of the WT-GWO-PLS calibration model was better than that of the WT-i PLS-VIP-PLS calibration model.The R~2p,RMSEP and MREP of the Lu element prediction set were0.9944,0.6276μg/g and 5.95%respectively.The R~2p,RMSEP and MREP of the Y element prediction set were 0.9932,56.8378μg/g and 5.81%respectively,and the relative errors(MRE)of rare earth elements Lu and Y were reduced by 21%and 24%respectively.Therefore,compared with i PLS-VIP hybrid variable selection method,using GWO as a variable selection strategy can improve the calculation efficiency and prediction accuracy of the model,and LIBS technology combined with WT-GWO-PLS can be used as an effective tool for accurate and quantitative analysis of rare earth elements Lu and Y in rare earth ores.(3)Integration of rare earth ore quantitative analysis system based on MATLAB App Designer platform can optimize the ore mining scheme by accurately determining the content of rare earth elements,and avoid ineffective mining and waste of resources.The system includes a sample set division module,spectral data preprocessing module,variable selection module and model construction module.By clicking the button,the button-related functions can be realized and the results of quantitative analysis can be displayed directly in the main interface of the system,and the forecast data can be saved.This system can basically meet the requirements of quantitative analysis of rare earth elements in rare earth ores,and can automatically and objectively obtain quantitative analysis results,and also provide some reference for other ore quantitative analysis processes.