【作者】 陈文菊；

【导师】 陈裕泉；

【作者基本信息】 浙江大学 ， 生物医学工程， 2005， 硕士

【摘要】 碳纳米管是理想的一维纳米材料,有许多独特的性质,其中其小尺寸、高的表面积及对气体的吸附特性使得它成为一种良好的气敏材料。本文主要研究基于碳纳米管的气敏传感器对有机气体的响应,并且尝试了胺基团化学改性碳纳米管和碳纳米管聚吡咯复合物等,观察其气敏影响,发现它们虽然对气敏响应的灵敏度改变不大,但是对苯、甲苯、二甲苯、甲醛的响应各有不同改变,为之后构造传感器阵列识别气体提供了基础。本论文也设计了基于各种碳纳米管传感器的电子鼻人工嗅觉系统识别有机气体如苯、甲苯、二甲苯、甲醛。采用主成分分析方法考察以上各基于碳纳米管传感器识别气体的可能性和降维处理,选择识别率较好的传感器构成阵列,构建径向基神经网络识别有机气体。 本论文首先介绍碳纳米管这种材料的结构、性质、制备及碳管吸附性的应用,尤其是其气敏方面的应用;然后介绍气敏传感器及气敏传感器阵列的种类、信号处理方法;之后是利用的各种修饰的碳纳米管制备的传感器进行气敏试验,对实验结果进行分析并对碳纳米管气体吸附的原理进行初步的探讨;然后根据之前各种修饰的碳纳米管的气敏传感器的气敏响应,通过PCA方法分析传感器阵列分类各有机气体的可能性,选择其中差异性大的构成传感器阵列,构造鲁棒性的神经网络,进行气体识别的模拟训练,并进行了简单的测试,在传感器阵列对气体的识别方面不受温度飘移和湿度影响。更多还原

【Abstract】 Carbon nanotubes are one-dimension nanomaterial with many particular properties. Their nano-size and high radio of surface proportion provide it to be promising gas sensitivity material. The paper mainly researched response of carbon nanotubes to volatile organic compounds (VOCs). The absorption of different gases on the carbon nanotubes layer changes the conductivity. We experimented on gas sensitive response of chemical modified carbon nanotube by adding amide-alkyl chain, doped with N elements and composite with poly-pyrrole. The gas sensitivity did not change significantly after the modification but were somewhat different, such as some are most sensitive to xylene; others are most sensitive to methanal.This paper also presented design of an electronic nose system based on gas sensors array and artificial neural network (ANN) for the identification of some (VOCs) relevant to environmental monitoring such as benzene, toluene; The sensors array consisted of the several modified carbon nanotubes-based sensors above. Principal component analysis (PCA) method was used for discussing the possibility of the sensors array for VOCs recognition and reduction of the dimensions for pattern recognition. Then designed the radial basis artificial (RBF) neural network by train and simulate the gas sensitivity data. A 100% accurate recognition ratio and a little error within 3 ppm by the interference of humidity were found by test experiment.The paper firstly introduced the structure, properties, preparation method and application of carbon nanotubes. Secondly electronic nose information was introduced. Thirdly the most important part was the gas sensitivity experiments. Next was the introduction of PCA and RBF ANN knowledge and then was the design work of the neural network including train, simulation and test. Finally the summarization and thanks were presented.更多还原