摘要
非法处理、携带与运输易燃易爆危险化学品(简称危化品)会对社会公共安全造成极大的威胁,能够对这些物质进行快速检出和鉴别可为危化品的监管与控制提供极大的便利。为了对易燃易爆危化品进行快速、准确的鉴别,提出了基于差示扫描量热法(DSC)对易燃易爆危化品进行鉴别的方法,并分析了测试条件对DSC测试结果的影响。结果表明:样品质量和温升速率对动态DSC谱图的影响与物质本身的物化特性有关,这为能够根据不同测试条件下的DSC谱图进行物质鉴别提供了可行性依据。同时,为了实现上述目的,从DSC谱图中提取了6个特征参数并建立了具有谱图分类识别功能的BP神经网络,选取15种易燃易爆危化品的DSC谱图数据对BP神经网络进行训练与测试,验证结果表明该BP神经网络对不同测试条件下的DSC谱图识别准确率较高,可为易燃易爆危化品的快速鉴别提供一种有效的方法。
Handling,carrying and transporting inflammable and explosive hazardous chemicals illegally would pose a great threat to social public security.A rapid detection and identification of those dangerous chemicals could provide great convenience for regulation and control.For the rapid detection and identification of inflammable and explosive hazardous chemicals,this paper proposes a new identification method for inflammable and explosive hazardous chemicals based on Differential Scanning Calorimetry(DSC)and investigates the influence of test conditions on DSC test.The results reveal that the influence of sample mass and heating rate is actually related to the physical and chemical properties of chemicals,which proves feasibility for chemicals identification with DSC spectrums under different test conditions.At the same time,in order to identify the inflammable and explosive hazardous chemicals with DSC spectrum without limiting the test conditions,the paper determines six characteristic parameters of DSC spectrum as a vector and builds a BP neural network with the function of spectrum pattern recognition.Then the paper uses DSC spectrum data of 15 kinds of inflammable and explosive chemicals to train and test the BP neural network.The validation result shows a high identification accuracy of DSC spectrums under different test conditions,which could prove potential application to identify the inflammable and explosive hazardous chemicals.
作者
王炎
陈利平
陈网桦
WANG Yan;CHEN Liping;CHEN Wanghua(School of Chemical Engineering,Nanjing University of Science and Technology,Nanjing 210094,China)
出处
《安全与环境工程》
CAS
北大核心
2018年第2期143-149,共7页
Safety and Environmental Engineering
基金
国家重点研发计划项目(2017YFC0804701-4)
关键词
易燃易爆危化品
鉴别方法
差示扫描量热法(DSC)
样品质量
温升速率BP神经网络
inflammable and explosive hazardous chemical
identification method
Differential Scanning Calorimetry(DSC)
sample mass
heating rate
BP neural network
