ZnO传感器与色谱分离相结合的POPs快速检测与识别方法

发布时间：2018-12-08 17:36

【摘要】：持久性有机污染物是人类合成的能持久的存在于自然环境之中,对人类及其他生物造成伤害的有毒化学物质。由于人类的大量的使用这类物质,导致了全球环境的污染。又因为持久性有机污染物具有持久性、生物积累性和广范围流动性,使得持久性有机污染物的在全球扩散和在生物体内累积,这些情况加重了持久性有机污染物对环境和人类危害。所以建立一个对持久性有机污染物的监测系统就显得尤为重要。有必要针对持久有机污染物对环境危害的实际情况,开发出持久有机污染物的检测设备,提高持久有机污染物检测的效率和准确性,为对持久有机污染物检测工作做出新的贡献。本文设计制作了一种可携带的持久性有机污染物检测仪器,可以对毒杀芬持久性有机污染物及干扰物进行检测。采用ZnO传感器与色谱分离相结合方法,结合计算机技术,达到对持久性有机污染物检测的目的,具有方便、快速、准确度高等特点。在仪器装置检测到的样本数据进行模式特征提取和训练识别前,我们对数据进行归一化处理,以消除因为量纲、样品浓度、环境温度等外界因素造成的影响。我们对数据进行特征提取时,采用PCA算法和LDA算法。最后根据提取的模式特征使用SVM和径向基人工神经网络训练出分类器,对待测样本进行分类识别。实验结果表明,该检测系统可以对持久性有机污染物进行分类识别,并有很好的分类效果。
[Abstract]:Persistent organic pollutants (pops) are toxic chemicals synthesised by human beings which can persist in the natural environment and cause harm to human beings and other organisms. The massive use of such substances by human beings has led to global environmental pollution. Also because persistent organic pollutants are persistent, bioaccumulative and mobile, leading to the global diffusion and accumulation of persistent organic pollutants in organisms, These conditions exacerbate the environmental and human hazards of persistent organic pollutants. Therefore, the establishment of a persistent organic pollutant monitoring system is particularly important. It is necessary to develop equipment for the detection of persistent organic pollutants in order to improve the efficiency and accuracy of the detection of persistent organic pollutants, in view of the actual situation of environmental hazards caused by persistent organic pollutants, To make a new contribution to the detection of persistent organic pollutants. In this paper, a portable detection instrument for persistent organic pollutants (pops) is designed and manufactured, which can be used to detect toxaphene persistent organic pollutants (pops) and interfering substances. The method of ZnO sensor combined with chromatographic separation, combined with computer technology, is used to detect persistent organic pollutants. It has the advantages of convenience, rapidity, high accuracy and so on. Before the pattern feature extraction and training recognition of the sample data detected by the instrument, we normalize the data to eliminate the influence caused by the external factors, such as dimension, sample concentration, environmental temperature, and so on. PCA algorithm and LDA algorithm are used in feature extraction of data. Finally, a classifier is trained by SVM and radial basis function (RBF) artificial neural network according to the extracted pattern features, and the samples are classified and identified. The experimental results show that the system can be used to classify and identify persistent organic pollutants and has a good classification effect.
【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：X830;TP212.9

【参考文献】