Study On Quantitative Analysis Method Of Heavy Metal In Soil Based On Laser-induced Breakdown Spectroscopy

The detection of heavy metal content in the soil is of great significance for soil quality assessment and pollution control.Laser-induced breakdown spectroscopy?LIBS?provides a simple and fast analysis method,which can achieve simultaneous measurement of multiple elements.Therefore,it is of great value to carry out quantitative analysis of heavy metal elements in soil based on LIBS technology.This article focuses on the influence of different pretreatment methods on the quantitative analysis of heavy metal elements,and establishes a quantitative analysis model of heavy metal element content in soil based on LIBS technology combined with chemometric methods,and analyzes the prediction performance.The main research contents of this article are as follows:?1?The construction of LIBS experimental platform.In this paper,a set of LIBS experimental platform is built based on the principle of LIBS and the spectral characteristics of soil samples.The experimental platform includes a host computer,laser,spectrometer,delay/pulse generator,sample stage,image acquisition device and optical path system,can realize the observation of the sample condition and the acquisition,display and storage of the spectral signal.?2?The influence of the LIBS experimental system parameters on the characteristic spectrum was studied,the obtained spectrum was preprocessed,and the abnormal spectrum was eliminated.Based on the National Institute of Standards and Technology database and the LIBS spectral information of the soil,the characteristic analysis lines of four elements Cu,Zn,Pb and Ni were selected.Through single factor analysis,the effects of LIBS experimental system parameters?laser pulse energy,delay time of the spectrometer received signal,laser frequency and laser focus position?on the spectral signal were discussed,and the instrument parameters used in the experiment were determined.The effects of normalization,Savitzky-Golay smoothing,and derivative pretreatment methods on the analysis results were compared,and the derivative was finally selected as the spectral pretreatment method in this study.Principal component analysis combined with Mahalanobis distance?PCA-MD?method was used to remove the abnormal spectral data of the preprocessed spectral data,and the input variables of the model were optimized.?3?Research on quantitative analysis methods of heavy metals in soil.With the help of the ability of the random forest algorithm to calculate and arrange the importance of each variable in the LIBS spectral data,filter the input variables according to the variable importance threshold,and construct a random forest based on variable importance?VI-RF?model.The two important parameters of the VI-RF correction model(n_tree:the number of decision trees and m_try:the number of randomly selected attributes)are optimized.After the optimization,the model's prediction results for the four elements Cu,Zn,Pb,and Ni have a root-mean-square error RMSE not greater than 12.6325,a decision coefficient R² not less than 0.9468,and a relative error not greater than 0.0660 of the prediction set;the root mean square error RMSE of the training set prediction result is not greater than 10.6892,the determination coefficient R² is not less than 0.9897,and the relative error is not greater than 0.0558.A partial least squares?PLS?quantitative analysis model was constructed with the same training set,and the performance of the PLS correction model was compared with the VI-RF model.The results showed that the VI-RF correction model had better quantitative accuracy.In order to study the stability of the prediction results of the VI-RF correction model,the robustness of the prediction results of the model is evaluated.The average RSD of the prediction results of the four elements to be tested are all less than 5.51%,which means that the VI-RF correction model has a good prediction stability.This paper confirms that the quantitative analysis of heavy metal content in soil based on LIBS technology combined with chemometrics is feasible,which lays the technical foundation for real-time monitoring of soil heavy metals in subsequent LIBS instruments.