PDF全文下载:2015050482
黄翼虎1, 栾浩杰1, 刘淑泉1, 张硕1, 马吉祥2, 林其安1
(1.青岛科技大学 自动化与电子工程学院,山东 青岛 266042; 2.山东泰开自动化有限公司,山东 泰安 271000)
摘要: 基于农药的太赫兹吸收光谱和支持向量机对4种农药进行了分类。利用太赫兹时域光谱系统对4种农药,即三氯杀螨砜、敌百虫、亚胺硫磷和六氯苯进行检测,获得样品的太赫兹时域信号,然后计算获得其在03~22 THz范围内的吸收光谱。用主成分分析法对太赫兹吸收光谱进行分析,提取光谱特征,用支持向量机分别对太赫兹吸收光谱和光谱特征进行分类。实验结果表明,这两种方案均能有效地对农药进行分类,分类正确率为100%。
关键词: 太赫兹吸收光谱; 农药; 主成分分析; 支持向量机
中图分类号:TS 255 文献标志码: A
Pesticide Classification Based on Support Vector Machine and Terahertz Spectroscopy
HUANG Yihu 1, LUAN Haojie1, LIU Shuquan1, ZHANG Shuo1, MA Jixiang2, LIN Qian 1
(1.College of Automation and Electronic Engineering, Qingdao University of Science and Technology,Qingdao 266042,China;
2.Shandong Taikai Automation Co.Ltd, Taian 271000, China)
Abstract: Based on pesticide terahertz absorption spectroscopy and support vector machines for four pesticides are classified. Use of terahertz timedomain spectroscopy system for four pesticides that tetradifon, trichlorfon, phosmet and hexachlorobenzene were detected to obtain a sample of terahertz timedomain signal, and then calculating gain its 03—22 THz absorption spectrum within the range. Principal component analysis of the THz absorption spectra were analyzed to extract the spectral characteristics of terahertz absorption spectra and spectral characteristics were using support vector machine for classification, the experimental results show that the two programs can effectively classify pesticides classification the correct rate of 100%.
Key words: terahertz absorption spectroscopy; pesticide; principal component analysis; SVM
收稿日期: 20140617
基金项目: 国家自然科学基金项目(61340038);山东省自然科学基金项目(2013ZRB01254);济南出入境检验检疫局项目(2012IK182).
作者简介: 黄翼虎 (1971—), 男,副教授. *通信联系人.
文章编号:16726987(2015)05048205; DOI: 10.16351/j.16726987.2015.05.002