离子迁移谱的智能识别法及应用于易制毒化学品稽查  

An intelligent recognition program for in-situ detection of illicit chemicals by ion mobility spectrometry

作　　者：陈新华[1] 姚琏[2] 李安林[2] 李京华[1] 刘广生[1] 何川先[1] 吴晓琳[1] 王海龙[1] 李海洋[1] 

机构地区：[1]中国科学院大连化学物理研究所,辽宁大连116023 [2]安阳工学院,河南安阳455000

出　　处：《计算机与应用化学》2007年第9期1145-1148,共4页Computers and Applied Chemistry

基　　金：国家重点基础研究发展规划(863)资助项目(2005AA641090)

摘　　要：重点介绍一台自主研制、用于易制毒化学品现场稽查的真空紫外灯电离-离子迁移谱仪,以及为这台仪器开发的一套数据处理和智能识别软件。数据处理包括数据采集、傅立叶变换和小波变换降噪等。智能识别方面建立了一套利用迁移时间匹配、夹角余弦和命中率指数3个参数作为综合评价指数,智能分析图谱的识别方法。该系统在对常见的易制毒化学品的检测中,识别率达到90%。Ion-mobility spectrometry has been known best for screening explosives at airports, detecting chemical agents for the military, and monitoring stack gas emissions in industry. This article generally introduces principle, character and developments of the instrumentation of ion mobility spectrometry. A homemade VUV-ion mobility spectrometer for in-situ detection of illicit chemicals is introduced. A software for data processing and intelligent recognition has been developed for this instrument. Fourier transform and wavelet transform methods are used to reduce noise in the data processing. A compound index method, which combines the matching rate of drift time, the angle cosine value and the hit quality index, is used for spectra analysis. This system has been tested in the detection of common illicit chemicals, producing a recognition rate of 90%.

关 键 词：离子迁移谱 模式识别 数据采集 

分 类 号：O6-64[理学—化学]