火灾早期过程特征的红外光谱探测和诊断  

DETECTION AND ESTIMATION OF EARLY FIRES' PROCESS CHARACTER BY USING INFRARED SPECTRUM

作　　者：张晓华[1] 张认成[1] 龚雪[1] 黄湘莹[1] 

机构地区：[1]华侨大学机电学院,福建泉州362021

出　　处：《红外与毫米波学报》2006年第5期382-385,共4页Journal of Infrared and Millimeter Waves

基　　金：福建省自然科学基金项目(D020015)

摘　　要：提出用CO作为早期火灾探测的观察对象,利用红外光谱法进行气体探测,并在此基础上建立了基于FTIR的试验系统.通过该方法成功获得CO的浓度值.并用时间序列分析方法建立二阶自回归模型,提取出特征参数进行分析,能在较短时间内将真假火灾区分开来.It was proposed that CO could be chosed as a target gas to detect early fires. A specific experiment system was set up based on FTIR technology. By using time series analysis, an auto-regressive models for the concentration of CO can be built. Finally, characteristic parameters of models can distinguish the real fire source from nuisance sources in short times.

关 键 词：红外光谱 吸收谱 时间序列分析 模型 

分 类 号：TP274.52[自动化与计算机技术—检测技术与自动化装置]