硝基漆品牌厂家红外光谱鉴别及预测模型建立  

Identification and prediction of nitro paints’brand manufacturers by infrared spectroscopy

作　　者：李德龙 王继芬[1] 刘项宇 黄涛[1] 陈浩然 马体 LI De-long;WANG Ji-fen;LIU Xiang-yu;HUANG Tao;CHEN Hao-ran;MA Ti(School of investigationand forensic science,People’s Public Security University of China,Beijing 102600,China)

机构地区：[1]中国人民公安大学侦查与刑事科学技术学院,北京102600

出　　处：《化学研究与应用》2020年第8期1443-1448,共6页Chemical Research and Application

摘　　要：在法庭科学领域,硝基漆的检验鉴别是一项重要的工作。为提高检验工作效率,提高分析可靠性,提出一种基于中红外光谱结合化学计量学的硝基漆鉴别方法。实验采集并获取长颈鹿等6种品牌共计59个样本的红外光谱数据,借助主成分分析、K近邻模型和判别分析构建分类模型。结果表明,PCA模型和K近邻模型对59个样本的区分能力相对较弱,判别分析模型的区分能力较强,实现了96.6%的准确区分和归类,分类结果理想。该方法能够快速,准确,无损地鉴别硝基漆,可为其他物证的检验鉴别提供一定的借鉴和参考。The identification of nitro paints is an essential task in the field of forensic science.In order to improve the efficiency and analysis reliability,a method for identifying nitro paints based on mid-infrared spectroscopy and chemometrics was proposed.Infrared spectral data of 59 samples from 6 brands including Giraffe were collected and obtained.A classification model was constructed by principal component analysis,K-nearest neighbor model and discriminant analysis.The differentiation of PCA and the K-nearest neighbor were relatively weak while the discriminant was stronger.96.6%of the samples were classified accurately,which was ideal.It can classify the nitro paints quickly,accurately and effectively and provide reference for other physical evidence.

关 键 词：硝基漆 中红外光谱 化学计量学 鉴别 

分 类 号：O657.33[理学—分析化学]