探讨纺织纤维的机器嗅觉快速识别方法  

Discussion on Rapid Identification Method of Textile Fiber by Machine Olfaction

作　　者：张晓利[1] 贾立锋[2] 梁家豪 巫莹柱[3] 孙运龙[2] ZHANG Xiaoli;JIA Lifeng;LIANG Jiahao;WU Yingzhu;SUN Yunlong(Guangdong Inspection and Quarantine Technology Center, Guangdong Guangzhou, 510623;Guangdong University of Technology, Guangdong Guangzhou, 510623;Wuyi University, Guangdong Jiangmen, 529020)

机构地区：[1]广东检验检疫技术中心 [2]广东工业大学 [3]五邑大学

出　　处：《棉纺织技术》2018年第6期26-29,共4页Cotton Textile Technology

基　　金：广东检验检疫局科技攻关项目(2016GDK43;2018GDK18);江门市科技攻关项目(江科[2016]189号);国家标准制修订项目(20172272-T-608);广东省教育厅特色创新项目(2016GXJK162;2017KTSCX182)

摘　　要：探讨机器嗅觉技术用于纤维鉴别的可行性。运用PEN3型电子鼻提取纤维材料挥发的气味信息,建立气味指纹图谱,选取传感器稳定后的合适试验数据,分别应用线性判别分析和主成分分析对腈纶、蚕丝、莫代尔、醋酯纤维、锦纶、氨纶的测试数据进行分析,并尝试将该技术用于棉涤混纺织物的检测中。结果表明:线性判别分析方法和主成分分析方法都能很好地识别所试验的纤维材料,其中线性判别分析方法所得数据点更为集中明显;PEN3型电子鼻能很好地分辨出待测棉涤混纺产品中的涤纶纤维和棉纤维。认为:该技术为快速、便捷、无损鉴别纤维种类提供了一种全新的方法。The feasibility of fiber identification by machine olfaction technology was discussed.PEN3 electronic nose was used to extract the odor information that was volatilized from fiber materials.The odor fingerprint was established.Appropriate experimental data was selected after sensor was stable.The test data for acrylic fiber,silk,modal,acetate fiber,polyamide fiber and polyurethane fiber were analyzed by adopting linear discriminant analysis and principal component analysis respectively.The technology was also used on the detection of cotton polyester blended fabric.The test results showed that both linear discriminant analysis and principal component analysis could well distinguish the tested fibers.Therein,the data points obtained by linear discriminant analysis were more obviously concentrated.Polyester fiber and cotton fiber in cotton polyester blended fabrics could be well distinguished by PEN3 electronic nose.It is considered that a brand new method is provided by the technology for quick,convenient and nondestructive fiber category identification.

关 键 词：机器嗅觉 纤维材料 快速鉴别 无损检验 模式分析 

分 类 号：TS107.2[轻工技术与工程—纺织工程]