分析化学Issue(7)：937-941,5.DOI:10.11895/j.issn.0253-3820.131136
用自组织特征映射神经网络对飞行时间质谱采集的大气气溶胶单粒子进行分类
Classification of Atmospheric Individual Aerosol Particles Sampled by Time-of-flight Mass Spectrometry Using Self-Organizing Map
摘要
Abstract
Large amount of data including chemical composition and size information of individual particles would be generated in the measurement of aerosol particles using atmospheric aerosol time-of-flight mass spectrometry ( ATOFMS ) . Our home-made ATOFMS was used to measure the indoor individual aerosol particles in real-time for 24 h, and the obtained mass spectrometric data were clustering analysis by self-organizing map ( SOM ) because of its ability of vector quantization and data dimensionality reduction. 20 classification results were got which included"Calcium-Containing","Salt+Secondary particles","Secondary particles","Organic Amines","K+-Rich Organics" and"Soil" particles, etc. Compared with previous mass spectrometric methods, SOM is a natural visualization tool, more classification results can be obtained. This classification information would be useful to assess the response and toxicity of atmospheric aerosol particles and identify the origin of atmospheric aerosol particles.关键词
气溶胶单粒子/气溶胶飞行时间质谱/自组织特征映射/聚类分析Key words
Individual aerosol particles/Aerosol time-of-flight mass spectrometry/Self-organizing map/Clustering analysis引用本文复制引用
郭晓勇,稳国柱,黄德双,方黎,张为俊..用自组织特征映射神经网络对飞行时间质谱采集的大气气溶胶单粒子进行分类[J].分析化学,2014,(7):937-941,5.基金项目
本文系国家自然科学基金(No.20477043)资助项目@@@@This work was supported by the National Natural Science Foundation of China (No.20477043) （No.20477043）
