鞋底材料的中红外光谱可视化快速鉴别  

Rapid Identification of Sole Materials with Mid-infrared Visible Spectroscopy

作　　者：王继芬 高春芳 徐佰祺 董泽 何欣龙 WANG Jifen;GAO Chunfang;XU Baiqi;DONG Ze;HE Xinlong(Institute of Forensic Science and Technology, People’s Public Security University of China, Beijing 102600, China;School of Data and Target Engineering, Strategic Support Force Information Engineering University, Zhengzhou 450000, China;School of Public Order Policing, People’s Public Security University of China, Beijing 102600, China)

机构地区：[1]中国人民公安大学刑事科学技术学院,北京102600 [2]战略支援部队信息工程大学数据与目标工程学院,郑州450000 [3]中国人民公安大学治安学院,北京102600

出　　处：《中国塑料》2019年第8期101-105,共5页China Plastics

基　　金：中国人民公安大学2019年度基本科研业务费项目(2019JKF332)

摘　　要：借助中红外分析技术和化学计量分析,建立常见鞋底材料种类鉴别的分类模型,为鞋底种类鉴别提供有效的新方法。基于鞋底红外光谱矩阵数据,建立了层次聚类模型和K近邻模型。结果表明,K近邻模型中,43个样本分为7类,这与层次聚类模型中,组间平均距离为8时样本的分类结果一致;K近邻下,分类样本呈5个聚集区域,其中H、G、K 3类样本分布情况明显,相比较A类,E类分布较为集中,聚敛程度较大,L类分在E类,M类分在A类;模型总分类准确率达95.35%,分类结果理想;选取各类中的一个样本,对其开展谱图解析,得到样本主要成分有顺丁橡胶、乙烯-醋酸乙烯酯共聚物、聚氨酯、异戊橡胶、聚氯乙烯和苯乙烯-丁二烯-苯乙烯嵌段共聚物;中红外光谱技术结合化学计量分析可作为鞋底材料种类鉴别的一种无损、快速、准确的分析方法。In this paper, a model for identifying the type of common sole materials was established by means of mid-infrared analysis technology and chemometrics. This model can provided an effective new method for identification of the material type of soles. A hierarchical cluster analysis was conducted and a K-nearest neighbor model was set up on the basis of the infrared spectral matrix data of soles. The results indicated that 43 samples could be divided into 7 categories in the K-nearest neighbor model, which is consistent with the categorizing result that the average distance between groups was 8 in the hierarchical clustering. Under the K-nearest neighbor condition, the samples were distributed in five aggregation regions, among which H, G and K samples exhibited an obvious distribution. Compared to the class A, the distribution of class E was relatively concentrated, and the degree of convergence was larger. The class L and M were split into the class E and A, respectively. The total identification accuracy of the model reached 95.35 %, indicating an ideal categorizing result. In addition, one sample of each category was selected and analyzed with spectrogram. The main components of them were achieved as butadiene rubber, ethylene-vinyl acetate copolymer, polyurethane, isoprene rubber, polyvinyl chloride and styrene-butadiene-styrene block copolymer. In summary, the mid-infrared analysis technology combined with chemometrics can realize a non-destructive, rapid and accurate identification for common sole materials.

关 键 词：鞋底材料 中红外分析 层次聚类 K近邻 鉴别 

分 类 号：TQ322[化学工程—合成树脂塑料工业]