摘要
【目的】研究利用便携式拉曼光谱仪测定农药氧乐果含量(质量分数)的定量分析方法,为农药残留的快速检测提供参考。【方法】利用便携式拉曼光谱仪,采集不同含量氧乐果溶液的拉曼光谱,比较不同预处理方法 (卷积平滑、一阶导数、二阶导数、基线校正、基线校正+卷积平滑)对氧乐果溶液拉曼光谱数据的预处理结果,从中筛选出最优预处理方法;以最优预处理方法处理之后的数据作为输入,分别采用偏最小二乘法(PLS)、主成分回归(PCR)和多元线性回归(MLR)算法建立拉曼光谱与氧乐果含量之间的预测模型,并对模型的预测效果进行评价。【结果】5种预处理方法中,氧乐果溶液拉曼光谱数据经基线校正处理后,所建立模型的预测效果最好。分别采用PLS、PCR和MLR算法建立了拉曼光谱与氧乐果含量之间的预测模型,对3个模型的预测结果进行评价,结果显示,PLS预测模型的相关系数(RP)为0.984,预测均方根误差(RMSEP)为1.12;PCR预测模型的RP为0.977,RMSEP为1.39;MLR预测模型的RP为0.539,RMSEP为5.54,可知基线校正与PLS结合建立的模型预测效果最佳。【结论】建立了基于便携式拉曼光谱仪的氧乐果残留的定量分析方法,为农药残留检测提供了一种快速、简单、方便的方法。
[Objective] This study quantitatively analyzed omethoate content using portable Raman spectrometer. [Method] The Raman spectroscopy of omethoate solutions with different mass fractions was collected using portable Raman spectrometer,and effects of different pretreatment methods （Savitzky Go- lay Smoothing, 1st derivatives, 2nd derivatives, base line, and base line＋ Savitzky Golay Smoothing） were investigated to select the optimal pretreatment method. Prediction models between Raman spectroscopy and mass fraction of omethoate were established and evaluated using partial least square （PLS）,principal com- ponent regression （PCR） and multiple linear regressions （MLR）. [Result] Among the 5 pretreatment methods, base line method had the best prediction results. Prediction correlation coefficients （Rv） of PLS, PCR, and MLR models were 0. 984,0. 977 ,and 0. 539, respectively. The values of root mean square error of prediction （RMSEP） for PLS, PCR, and MLR were 1.12,1.39, and 5.54, respectively. Combination of base line pretreatment method and PLS had the best results. [Conclusion] A quantitative analysis method of omethoate residues based on portable Raman spectrometer was established, which was fast, simple and con venience for detection of pesticide residues.
出处
《西北农林科技大学学报（自然科学版）》
CSCD
北大核心
2014年第2期136-141,共6页
Journal of Northwest A&F University(Natural Science Edition)
基金
国家自然科学基金项目(31160250,61178036)
江西省主要学科学术和技术带头人培养对象计划项目(2009DD00 700)
赣鄱英才“555”工程领军人才培养计划项目(2011-64)
江西省光电检测工程技术研究中心资助项目(赣科发财字[2012]155号)
