近红外光谱技术用于桑叶中毒死蜱残留的定量分析研究

摘要本研究提出了应用近红外光谱技术对桑叶中的毒死蜱残留进行定量分析的研究策略。运用近红外光透射模式来检测农药毒死蜱，检测的近红外光光谱波长范围在 1250-1700nm, 采用分辨率为 8cm-1，采样次数 32 次。为了模拟果疏生长的自然背景，前期配制了 39 个有机磷农药毒死蜱浓度为 0。005～0。1mg/kg 的模拟混合溶液样本进行预实验，以确定其可行性；后期对浸泡在毒死蜱浓度为 0。005～0。1mg/kg ，浓度梯度为 0。0025mg/kg 的 39 个桑叶样品进行定量分析研究。随机挑选 26 个样品作为校正集，余下的样品作为预测集。经过对光谱预处理数据的比较，得出二阶导数结合正态变量变化是最佳光谱预处理方法。然后，通过选择光谱富集波段以及挑选出最适宜的主因子数，建立起基于偏最小二乘回归算法的近红外光谱的数学模型。建立模型校正集中的样本的预计值与实际值之间的相关系数为 0。9844，预测集中的样本预测值与实际值之间的相关系数为 0。9445，交叉验证均方根 RMSECV 为 0。0075，预测均方根 RMSEP 为 0。0067，证明了研究获得的模型具有较好的预测能力和较高的预测精准度。 83573

毕业论文关键词：近红外光谱技术；毒死蜱；偏最小二乘法；定量分析研究

Abstract This study proposed the research strategy on quantitative analysis of chlorpyrifos residues in mulberry leaves by using near infrared technology。 Using near infrared transmission mode to detect Chlorpyrifos, detection of near infrared spectral wavelength range in 1250-1700nm, with a resolution of 8cm-1, sampling times 32 times。 In order to simulate the natural background of fruit thinning, early made up 39 organophosphorus pesticide chlorpyrifos concentrations ranging from 0。005 to 0。1mg/kg analog mixed solution sample to conduct Pre experiment, to determine the feasibility; Late using the 39 samples from mulberry leaves which soaked in chlorpyrifos concentrations in 0。005~0。1mg/kg, the concentration gradient of 0。0025mg/kg to make quantitative studies。 26 samples were randomly selected as the calibration set and the remaining samples were used as predictor。 Comparison of spectral pretreatment data, it is concluded that the two order derivative combined with the normal variable is the best method of spectral pretreatment。 Then, selected spectral enrichment band and the most appropriate number of main factors, a mathematical model of near infrared spectroscopy based on partial least squares regression was established。 The correlation coefficient between the estimated value and the actual value

of the model is 0。9844，the correlation coefficient between the sample forecast value and

the actual value is 0。9445, cross validation root mean square RMSECV is 0。0075 and root mean square root mean square RMSEP is 0。0067, It was proved that the model has good predictive ability and higher precision。

Keywords: near infrared；chlorpyrifos；partial least square method ; quantitative analysis of the research；

第一章绪论。 1

1。1 引言 1

1。2 研究背景。 1

1。3 桑叶毒死蜱残留的研究意义。 2

1。3。1 桑叶的价值 2

1。3。2 桑叶衍生物在世界范围内的影响 3

1。4 毒死蜱的益弊。 4

1。4。1 毒死蜱简介 4

1。4。2 毒死蜱的危害性 4

1。5 近红外光谱技术的概述。 5

1。5。1 现代近红外光谱技术发展历程 5

1。5。2 近红外光谱法的基本原理 6

1。5。3 现代近红外光谱技术的特点 6

1。5。4 近红外光谱定量校正模型 7 近红外光谱技术用于桑叶中毒死蜱残留的定量分析研究:http://www.youerw.com/shengwu/lunwen_98625.html