摘要
为了对现场大量缴获毒品进行快速检测,实现无损准确定性分析,实验对神经兴奋剂、神经抑制剂和致幻剂三类毒品共166份样本的红外光谱数据进行采集,利用径向基函数神经网络与支持向量机构建不同数据分类模型,并对比基于不同核函数的支持向量机模型对分析准确率的影响。结果表明,在三类毒品样本识别与分类过程中,基于Polynomial核函数的支持向量机分类模型分类效果优于其他模型,样本训练集的分类准确率达到96.5%。该模型在对同一类型中不同种类毒品样本的识别与分类过程中,神经兴奋剂类中各种类毒品样本训练集的分类准确率达到96.4%。本研究实现了不同类型及同一类型不同种类毒品快速准确的定性分析,为这类走私管制类药物案件的准确定性及刻画犯罪嫌疑人相关行为特征提供了一定的技术支持。
In order to quickly inspect the drugs seized at the scene and achieve non-destructive and accurate qualitative analysis,the experiment collected the infrared spectral data of 166 samples of three types of drugs,central nervous depressants,central nervous stimulants and hallucinogens,use radial basis function neural network and support vector machine to build different data classification models,compare the impact of support vector machine models based on different kernel functions on the accuracy of the analysis.The results show that in the process of identifying and classifying three types of drug samples,the support vector machine classification model based on the polynomial kernel function is better than other models,and the classification accuracy of the sample training set reaches 96.5%.In the process of identifying and classifying different kinds of drug samples in the same type,the classification accuracy of the training set of various types of drug samples in the central nervous stimulants class reached 96.4%.The research has realized the rapid and accurate qualitative analysis of different types and different kinds of drugs of the same type,and provided clues and evidence for accurately identifying the source of seized drugs and trying the upstream and downstream drug crime cases related to facts.
作者
李佳瑞
王继芬
石学军
郭宇轩
LI Jia-rui;WANG Ji-fen;SHI Xue-jun;GUO Yu-xuan(School of Investigation,People's Public Security University of China,Beijing 100038,China;Forensic Expertise Center of Beijing Customs Anti-smuggling Bureau,Beijing 100000,China)
出处
《化学研究与应用》
CAS
CSCD
北大核心
2022年第7期1517-1525,共9页
Chemical Research and Application
关键词
光谱分析
神经兴奋剂
神经抑制剂
致幻剂
径向基函数
支持向量机
spectral analysis
central nervous depressants
central nervous stimulants
Hallucinogens
radial basis function
support vector machine
