【摘要】 分析了傅里叶变换红外光谱仪（FTIS）遥测的目标气体和几种干扰物的红外光谱特征，给出红外光谱的模式识别方法用于远距离有毒气体环境的监测。用选定的标准样本进行训练，得到稳定的权值和阈值。该方法适合于对局部环境内有毒、有害气体成份的远距离实时监测。具有较高的识别率。应用的结果表明：该方法对特定成份的局部环境监测是十分有效的。更多还原

【Abstract】 The spectral features of target-gas and several interfences by Fourier transform infrared spectrometer (FTIS) are analysed. Infrared spectral pattern-recognition methods are applied to the environmental monitoring. Selected standard samples are trained, the stable weight values and threshold value of the system are obtained. This method is suitable for real-time remote detection of hazardous-gas in local environment. It shows very high discrimination results. The results of its practical examples for local atmospheric environment monitoring prove that pattern recognition methods are quite effective for the fast discrimination specific component of environmental monitoring.更多还原