早期火灾CO气体的时间序列建模

Time Series Modeling of CO in the Early Fire Detection

通过傅里叶变换红外光谱探测技术,对火灾燃烧过程中产生的CO气体体积分数进行探测和定量分析,并对现场实测数据的非平稳时间序列进行平稳化处理和AR(p)模型(自回归模型)建模.通过对模型参数在相平面上的分析,提取早期火灾发生过程中的特征信息,建立一种及时的火灾探测、报警的方法.

The time series analysis method has been widely used and succeeded in a lot of fields of engineering. In this paper, we detect the volume fraction of CO in the combustion by the theory of FTIR （Fourier Transform Infrared） spectroscopy, and nonstationary time series method is used in the AR（p） model （auto-regressivemodel） distinction. It＇s feasible to reduce the false alarms and increase sensitivity of early fire detection through the identificating and modeling of the concentration of the gas.