基于比色光谱和SVM快速检测白菜中氧乐果农残方法

为快速安全地检测白菜中氧乐果农药残留,用改进的氯化钯乙酸溶液作比色剂,和氧乐果农药发生比色反应,利用分光光度计在300～1100nm采集吸光度谱图。将氧乐果乳油农药和白菜汁混合,利用乙醇萃取,在0.5～400mg/kg随机配置了60个样本,分别和等量的乙酸氯化钯溶液反应。实验证明,当样本质量比大于50mg/kg时,吸光度曲线的走势发生变化,且和质量比的相关性较差；当样本质量比大于90mg/kg时,吸光度曲线和样本质量比相关性更差。引入支持向量机SVM进行三分类建模,利用5-折交叉验证,当光谱数据归一化为0,1],核函数采用径向基RBF函数时,测试集样本质量比范围的预测效果最好,准确率达到96.7742%,为快速检测蔬菜中氧乐果的质量比范围提供了可行的方法。

Rapid Detection of Omethoate Pesticide Residues in Chinese Cabbage Based on Colorimetric Spectroscopy and SVM

In order to fast and safely detect omethoate pesticide residues in Chinese cabbage, PdCl₂ dissolved in acetic acid was used as the colorimetric reagent to have colorimetric reaction with omethoate and the region of 300-1100nm was applied. The mixture of omethoate pesticide and Chinese cabbage juice was extracted using ethanol. The results of 60 samples with the concentration between 0.5mg/kg and 400mg/kg showed that shape of absorbance spectra was changed and a poor correlation between the absorbance and concentration was obtained when the concentration was higher than 50mg/kg. When the concentration was higher than 90mg/kg, a worse correlation between the absorbance and concentration was obtained. Three classification modeling was established by support vector machine (SVM). The classification accuracy of the validation set by 5-fold cross validation was up to 96.7742% when the absorbance data were normalized to 0,1] and radial basis function (RBF) was used as kernel function. This method provided a feasible method for rapid qualitative detection of omethoate concentration.