声表面波电子鼻探测器的技术研究

发布时间：2018-01-23 14:25

本文关键词： 电子鼻毒剂检测 SAW传感器阵列模式识别 PIC　出处：《电子科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：自第一次世界大战以来,化学战剂(CWA)大量被用于战场,近二三十年来,更被恐怖组织所利用,已经严重危害到生态环境的持续发展和全人类的生命财产安全。由于化学战剂具有易制造、造价低、杀伤力强、破坏范围广等特点,使得禁止化学战剂使用的国际公约至今还未成功禁止一些国家与组织对化学战剂的保有与研制。因此需要一种响应快、灵敏度高、操作简单的手持式实时检测仪器,以声表面波(SAW)传感器阵列为基础的痕量气体电子鼻具有体积小、响应快、易于携带等优点,比其它检测手段性能更优越。本课题就是在此背景下产生,利用涂覆有选择性吸附敏感膜的声表面波器件制备成SAW传感器阵列,对不同种类、不同浓度的气体进行气敏测试与分析,结合模式识别算法,最终可在dsPIC30F6014A单片机上实现对神经毒剂模拟剂(DMMP)、芥子气模拟剂(2-CEES与DCP)和干扰气体(丙酮、二氯甲烷、甲醇、乙醇)的识别,并在LCD上显示毒剂的种类与浓度。本文介绍了课题的意义、SAW气体传感器原理、SAW传感器阵列的测试平台、模式识别算法及单片机平台算法的实现,主要研究内容有以下几个方面:(1)声表面波传感器阵列的设计:介绍了以双通道SAW谐振器混频电路为基础的传感器阵列的功能结构;利用气喷工艺在3个SAW谐振器上分别喷涂氟多元醇(FPOL)、聚氰丙甲基硅氧烷(PCPMS)和聚环氧氯丙烷(PECH)这三种敏感膜,并与开盖SAW参比器件组成传感器阵列。(2)传感器阵列的气敏测试与分析:对三种毒剂模拟剂和四种干扰气体分别在10~300 mg/m3范围内进行系统测试和分析。结果表明FPOL-SAW对神经毒剂具有选择性,PECH-SAW对芥子气有选择性。(3)模式识别分析:利用SPSS统计软件对采集的数据分别用主成分分析法(PCA)和人工神经网络(ANN)算法对气体进行模式识别。主成分分析模型简单,可以用图表直观表示,人工神经网络算法易于理解,这两种算法都可以将毒剂模拟剂与干扰气体进行区分。(4)基于dsPIC30F6014A单片机的电子鼻系统算法实现:在单片机平台上实现电路控制与模式识别算法,文章分别实现了定时\计数、预测干扰信号、气体响应检测、气体识别功能、串口发送和液晶显示这六个模块,最后进行算法验证。
[Abstract]:Since World War I, the chemical warfare agent CWAs have been used in large numbers on the battlefield, and in the last 20 or 30 years, they have been used by terrorist organizations. Chemical warfare agents are easy to manufacture, low cost, strong killing power, wide range of destruction and so on. So far, the international convention against the use of chemical warfare agents has not been successful in prohibiting the retention and development of chemical warfare agents in some countries and organizations. Therefore, a rapid response and high sensitivity are required. The electronic nose of trace gas based on saw sensor array has the advantages of small volume, fast response and easy to carry. The performance of SAW sensor array is better than that of other detection methods. In this paper, the surface acoustic wave devices coated with selective adsorption sensitive film are used to fabricate the SAW sensor array, for different kinds of sensors. Gas sensitivity test and analysis of different concentrations of gas, combined with the pattern recognition algorithm, can be implemented on the dsPIC30F6014A microcontroller to simulate the nerve agent (DMMP). The identification of mustard gas simulants (LCD) and interference gases (acetone, dichloromethane, methanol, ethanol), and the types and concentrations of the agents on LCD. The significance of the subject is introduced in this paper. The principle of SAW gas sensor is used to realize the test platform of saw sensor array, the algorithm of pattern recognition and the algorithm of MCU platform. The main research contents are as follows: 1) the design of saw sensor array: the functional structure of sensor array based on dual-channel SAW resonator mixing circuit is introduced. Three kinds of sensitive films were sprayed on three SAW resonators by gas injection, including fluorine polyol, poly (cyanopropylmethyl) siloxane (PCPMS) and polyepichlorohydrin (SAW). The sensor array is composed of an open-cap SAW reference device. Gas sensitivity Test and Analysis of Sensor Array: three toxic simulants and four interference gases were measured at 10: 300. The results of systematic test and analysis within mg/m3 showed that FPOL-SAW was selective to nerve agent. PECH-SAW pattern recognition analysis for mustard gas: using SPSS software to analyze the collected data by principal component analysis (PCA) and artificial neural network (Ann). The principal component analysis (PCA) model is simple. The artificial neural network algorithm is easy to understand. Both algorithms can distinguish the toxic analogue from the interfering gas. The algorithm of electronic nose system based on dsPIC30F6014A microcontroller is realized: the circuit control and pattern recognition algorithm are realized on the platform of single chip microcomputer. In this paper, the six modules of timing\ counting, predicting interference signal, gas response detection, gas recognition function, serial port sending and liquid crystal display are implemented respectively. Finally, the algorithm is verified.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN65

【参考文献】