摘要
针对传统取样分析技术会破坏物证和综合考察样本作为混合物在多维度上的差异性,提出一种基于二阶导数红外光谱结合模式识别对轮胎橡胶颗粒快速准确鉴别的方法。采集并分析不同品牌共计96个轮胎橡胶颗粒的红外谱图及其二阶导数谱图,同时预处理采用自动基线校正、峰面积归一化和Savitzky-Golay平滑,建立判别分析模型,从而实现其品牌间的准确区分和认定。红外二阶导数谱图呈现出许多原始谱图中被掩盖谱峰的斜率变化特征,将样本谱图间的差异更为明显的表示了出来,结合原始谱图和其二阶导数谱图,得出实验样本主要由丁苯橡胶、顺丁橡胶和异戊橡胶3种类型。原始谱图判别预测模型分类准确率为95.83%,二阶导数判别预测模型分类正确率为100%,其区分能力更强,二阶导数结合判别分析可有效开展对轮胎橡胶颗粒的区分鉴别,其构建的模型分类效果更好。以品牌为单位,进一步对丁苯橡胶等3种类型的样本展开模式识别工作,得出其判别预测模型均实现了样本品牌间100%的区分和归类,实验结果理想。利用二阶导数红外光谱结合模式识别可实现对轮胎橡胶样本的准确识别与分类,方法具有一定的普适性和借鉴意义,可为其他物证的鉴别与分析提供一定的参考。
A method,based on second derivative IR spectra and pattern recognition,is proposed to nondestructively and quickly identify the tire rubber particles,aim at the problem of traditional analysis would destroy the evidence and the comprehensive consideration of multi-variable and multi-dimensional differences in samples.The paper collected and analyzed the infrared spectra and their second derivative infrared spectra of 96 tire rubber particles are from different brands,pre-processing used automatic baseline correction,peak area normalization and Savitzky-Golay smoothing.Finally,a discriminant analysis identification model was established to achieve an accurate identify.The infrared second derivative spectra showed the masked peaks slope variation characteristics in many original spectra and the clearer difference between the sample spectra.Combined with the original spectra and its second derivative spectra,the samples were mainly composed of styrene-butadiene rubber,butadiene rubber and isoprene rubber.The accuracy of the original spectral discriminant prediction model is 95.83%,and the second derivative discriminant prediction model is 100%,which is more distinguishable and can effectively identify the rubber particles of tires.Each type discriminant prediction model can also achieve 100%discrimination.In summary,second derivative infrared spectra combined with pattern recognition enables accurately to identify the tire rubber particles,which has universality and certain reference significance and can provide some reference for the identification and analysis of other physical evidence.
作者
何欣龙
王继芬
王飞
兰薪康
罗鑫
HE Xinlong;WANG Jifen;WANG Fei;LAN Xinkang;LUO Xin(Institute of Forensic Science and Technology,People’s Public Security University of China,Beijing 102600,China;School of Electronic Engineering,Naval University of Engineering,Wuhan 430000,China)
出处
《中国测试》
CAS
北大核心
2019年第9期60-64,83,共6页
China Measurement & Test
基金
中国人民公安大学2019年拔尖人才培养专项资助硕士研究生科研创新项目(2019ssky003)
关键词
轮胎橡胶颗粒
二阶导数红外光谱
判别分析
鉴别
tire rubber particles
second derivative infrared spectra
discriminant analysis
identification
