摘要
采用激光诱导击穿光谱法(LIBS)对婴幼儿配方奶粉中的钙元素进行检测,但由于Stark效应的干扰,其LIBS谱峰信号极易产生漂移和拓宽等畸变信息.为了有效克服LIBS信号畸变对定量分析的影响,结合了改进随机蛙跳算法(MRFA)、高密度小波变换(HDWT)和偏最小二乘法(PLS),以近似平移不变的方式选择LIBS信号的最佳特征,进而在LIBS信号畸变的情况下获得良好的定量效果.结果表明,MRFA-HDWT-PLS模型的预测均方根误差为0.029,相关系数0.97,相对均方根误差为5.3%,有效满足配方奶粉中钙元素检测的精度要求,有望为婴幼儿奶粉质量安全筛查提供一种新手段.
The laser induce breakdown spectroscopy(LIBS)was used to detect the calcium content in infant formula.However,the Stark effect presented in the LIBS often distorted the signals with shifted and broadened results.In order to effectively suppress the influence of LIBS signal distortion on quantitative analysis,the improved random frog algorithm(MRFA),high density wavelet transform(HDWT)and partial least squares(PLS)are combined to optimal features of LIBS signal in an approximate shift-invariant way.It is thus expected to obtain a good quantitative effect in the presence of distorted LIBS signals.As a result,the root mean square error of the MRFA-HDWT-PLS model is 0.029,the correlation coefficient is 0.97 and the average root mean square error is 5.3%.The result illustrates that the MRFA-HD-WT-PLS presents a promising tool for determination of calcium content in infant formula samples.
作者
陈达
刘俊鑫
张翠
马艺闻
黄志轩
Chen Da;Liu Junxin;Zhang Cui;Ma Yiwen;Huang Zhixuan(School of Precision bistmment and Opto-Electronics Engineering,Tianjin Universityy Tianjin 300072,China)
出处
《南开大学学报（自然科学版）》
CAS
CSCD
北大核心
2021年第2期80-84,共5页
Acta Scientiarum Naturalium Universitatis Nankaiensis
基金
国家自然科学基金面上项目(21973111)。
