论文部分内容阅读
激光诱导击穿光谱(LIBS)技术结合支持向量机(SVM)定量分析土壤中Cr元素的含量。利用波长为1 064nm的Nd∶YAG脉冲激光器作为激发光源,采用光栅光谱仪和CCD分光探测不同重金属元素含量土壤样品的LIBS特征光谱。为了提高土壤中Cr元素定量分析的精度,分别采用多元线性回归分析和SVM两种方法对土壤中Cr元素的含量进行定量分析。研究结果表明,采用多元线性回归分析方法可以有效提高定量分析的精度,定标曲线拟合相关系数从传统定量分析方法的0.689提高到0.980;SVM定量分析方法训练集得到的定标曲线斜率近似为1,拟合相关系数为0.998,优于传统定量分析方法和多元线性回归分析方法,对检验集的预测相对误差均在2.57%以内。LIBS技术结合多元线性回归和SVM定量分析方法可以有效的提高土壤中Cr元素定量分析的稳定性和精度,校正土壤基体效应对Cr元素定量分析的影响。
Laser Induced Breakdown Spectroscopy (LIBS) combined with Support Vector Machine (SVM) for the quantitative analysis of Cr in soil. The LIBS spectra of soil samples with different heavy metal contents were detected by grating spectroscopy and CCD spectroscopy using Nd: YAG pulsed laser with a wavelength of 1 064 nm as the excitation light source. In order to improve the accuracy of quantitative analysis of Cr in soil, the contents of Cr in soil were quantitatively analyzed by multiple linear regression analysis and SVM respectively. The results show that using multivariate linear regression analysis method can effectively improve the accuracy of quantitative analysis. The correlation coefficient of calibration curve fitting is increased from 0.689 to 0.980 by traditional quantitative analysis method. The slope of the calibration curve obtained by SVM quantitative analysis method is approximately 1, the fitting correlation coefficient is 0.998, which is better than the traditional quantitative analysis method and multivariate linear regression analysis method. The relative error of the test set is within 2.57%. LIBS combined with multiple linear regression and SVM quantitative analysis method can effectively improve the stability and precision of quantitative analysis of Cr in soil and correct the influence of soil matrix effect on quantitative analysis of Cr element.