摘要
利用优化的试验参数采集青菜样品中Cd元素的LIBS特征光谱,结合单变量定标法和偏最小二乘法(PLS)对青菜中的Cd元素进行定量分析。结果显示:基于样条(Spline)基线校正后的单变量定标法和PLS均能定量地预测青菜中Cd元素含量,且PLS定标模型性能更优,其定标集和预测集决定系数(R^2)和均方根误差分别为0.9690,0.8317mg/kg和0.9960,0.4089mg/kg,预测平均相对误差为2.56%。
Cd dement in green vegetables was studied by using the optimized parameters combined with single variable calibration and partial least squares (PLS).Results showed that the baseline correction (spline)after the univariate calibration method and PLS can quantitatively predict the content of Cd element in vegetables,and the PLS calibration model has better performance.The determination coefficient (R^2)and root-mean-square error of its calibration set and prediction set were 0.9690,0.8317 mg/kg and 0.9960,0.4089 mg/kg,respectively.The relative error of prediction was 2.56%.
作者
罗子奕
黄林
刘木华
陈添兵
陈金印
饶刚福
许方豪
何秀文
周华茂
林金龙
姚明印
LUO Zi-yi;HUANG Lin;LIU Mu-hua;CHEN Tian-bing;CHEN Jin-yin;RAO Gang-fu;XU Fang-hao;HE Xiu-wen;ZHOU Hua-mao;LIN Jin-long;YAO Ming-yin(College of Engineering,Jiangxi Agricultural University,Nanchang 330045;Collaborative Innovation Center of Postharvest Key Technology and Quality Safety of Fruits and Vegetables in Jiangxi Province,Nanchang 330045;Jiangxi Key Laboratory of Modern Agricultural Equipment,Nanehang 330045)
出处
《分析试验室》
CAS
CSCD
北大核心
2018年第12期1384-1388,共5页
Chinese Journal of Analysis Laboratory
基金
国家自然科学基金(31460419,31560482)资助
关键词
激光诱导击穿光谱
重金属CD
偏最小二乘法
定量分析
Laser induced breakdown spectroscopy (LIBS)
Heavy metal Cd
Partial least squares (PLS)
Quantitative analysis
