Research On The Method Of Improving The Detection Accuracy Of Heavy Metals In Soil Based On LIBS

Soil heavy metal pollution is one of the environmental problems facing China.Improving the detection accuracy of heavy metal element in soil and achieving high-precision and fast detection of heavy metal elements in soil is an urgent problem to be solved.Most existing soil heavy metal detection methods require complex sampling and sample preparation processes,which are time-consuming and cannot achieve fast and efficient multi-element detection.Laser-Induced Breakdown Spectroscopy(LIBS)is an emerging element detection technology that has developed rapidly in recent years due to its simple sample preparation and the ability to simultaneously detect multiple elements.However,the detection accuracy of LIBS technology in heavy metal element detection in soil is still not high enough.This is mainly due to the redundancy of LIBS spectral data,the complexity of soil composition,and the susceptibility of plasma spectra to sample matrix effects and self-absorption effects.In order to improve the detection accuracy of LIBS technology in soil heavy metal detection,this paper conducts research on LIBS detection device construction,feature variable selection,matrix effect and self-absorption effect correction,and other aspects.The specific research contents are as follows:(1)Design and construction of a LIBS detection deviceA LIBS detection device was built and the laser energy,focusing point position,and acquisition delay time were optimized.Based on this,an auxiliary focusing system was designed to assist in adjusting the sample excitation point position by fixing two laser pointers at a specific angle,reducing experimental errors caused by uneven sample thickness.Additionally,a signal enhancement device was designed to collect radiation photons from different spatial angles,coupling more spectral signals into the spectrometer for spectral analysis,thus improving the intensity and stability of the spectral signals.(2)Research on data preprocessing methods based on feature variable selectionIn order to solve the problem that mutual interference between spectral lines in univariate regression analysis,multivariate regression analysis method was used to improve the quantitative analysis accuracy.However,LIBS spectra have the characteristic of a large amount of data with few samples,and using full spectra data as the input of multivariate regression analysis is prone to overfitting.Therefore,the ALASSO feature variable selection algorithm was used to reduce the dimensionality of full spectral data before using LIBS spectral data for multivariate regression analysis.According to the contribution degree of different feature variables to quantitative regression,important related feature variables were selected for multivariate regression analysis,irrelevant feature variables were eliminated to improve the accuracy and precision of quantitative analysis,and reduce the risk of overfitting.(3)Research on the influence of soil matrix effect on quantitative analysis and compensation methodDue to the different soil matrix(composition,element content,soil particle size,moisture content,etc.),the physical and chemical properties between the standard sample and the test sample are different,which affects the accuracy of quantitative analysis.Taking soil sample surface hardness as an example,this paper studies the effect of different surface hardness on plasma temperature,ion lines to atomic lines ratio,and calibration curve,to analyze the influence of soil matrix on detection accuracy.By comparing the ability of PLSR and LSSVM multivariate regression models to handle sample matrix effects,the interference of different soil matrix on quantitative analysis was compensated in the data processing stage to improve the performance of quantitative analysis.(4)Research on the influence of plasma self-absorption effect on quantitative analysis and compensation methodConsidering that the calibration curve exhibits a saturation phenomenon due to the self-absorption effect as the concentration of the detected element increases,this paper studies the nonlinear effect of plasma self-absorption effect on the calibration curve,and evaluates the self-absorption coefficient of each spectral line.On this basis,a segmented fitting method and SA-LSSVM algorithm were proposed to correct the nonlinear effect of plasma self-absorption effect on quantitative analysis,achieving high-precision detection of heavy metal elements in soil.This paper achieves high-precision detection of heavy metal elements in soil,and improves the detection accuracy of LIBS soil heavy metal detection from four aspects: the design of the LIBS detection device,feature variable selection,matrix effect and self-absorption effect correction of sample.This further promotes the commercialization process of LIBS detection technology in the field of soil heavy metal detection.