改进的谱峰拟合光谱标准化方法定量分析废钢中元素

Quantitative Analysis of Elements in Scrap Steel by Improved Peak Fitting Spectral Standardization Method

为了解决因设备和环境因素导致激光诱导击穿光谱(LIBS)技术在定量测量时的较高波动性问题,提出了基于改进的沃伊特函数拟合的光谱标准化方法。首先通过高精度激光器采集废钢样品的LIBS光谱数据,然后通过标准等离子体条件(标准等离子温度、电子数密度、待测元素总密度)得到标准光谱强度与实际谱线强度、半峰全宽等因素之间的关系式,并基于近似沃伊特函数对相关的谱峰进行拟合,获得标准化的重要参数,完成对样品的标准化模型的建立。结果表明,将该标准化方法用于废钢中铜、镍、硅、铬、锰元素的定量预测,得到的相关系数(R^(2))、预测均方根误差(RMSEP)和平均相对误差(ARE)的平均值分别为0.9814,0.063,6.3%,有效降低了波动影响,增强了分析精度,能够在工业中对废钢元素进行快速检测。

In order to solve the problem of higher volatility of laser induced breakdown spectroscopy(LIBS)in quantitative measurement due to equipment and environmental factors,a spectral standardization method based on improved Voigt fuction fitting was proposed.First,the LIBS spectrum data of the scrap steel sample was collected with a high-precision laser;then the relationship among the factors including the standard spectrum intensity and the actual spectrum line intensity,full width at half maximum,etc.,was obtained through standard plasma conditions(standard plasma temperature,electron number density,total density of elements to be tested).The approximate Voigt function was used for fitting the relevant peaks,obtaining the important parameters of the standardization,and the process of establishing the standardization model of the samples was completed.As shown by the results,this standardized method was used for quantitatively predicting the contents of elements(Cu,Ni,Si,Cr,Mn)in scrap steel,and the average values of correlation coefficient of determination(R^(2)),predicted root mean square error(RMSEP),and the average relative error(ARE)were 0.9814,0.063,and 6.3%,respectively.This method effectively reduced the impact of fluctuations and enhanced the accuracy of analysis,which could quickly detect the elements in scrap steel in industry.