基于小波分解的油料火焰光谱特性分析研究

Analytic Study on Spectrum Characteristics of Oil Flame Based on Wavelet Decomposition

针对油料火焰光谱特性研究不足的现状，通过构建全火焰红外测试系统，在室外开放空间条件下对多种油料及其他可燃物火焰的发射光谱进行了测试分析研究，光谱范围为1—14μm。结果表明，蜂窝煤原始光谱信号最为特殊。其他燃料火焰光谱信号经db2小波5层分解后，92#汽油、95#汽油、0#柴油、航空煤油、润滑油火焰光谱低频分量特征相似，在1．2、3．4、4．5gm附近存在较强的发射峰。各油料火焰光谱第5层细节系数重叠度较高。92#汽油、0#柴油火焰光谱低频分量及细节系数与其他燃料（木柴、酒精及纸张）相比特征明显。实验结论对基于光谱特性分析的油料火焰探测识别具有重要借鉴意义。

Due to the lack of adequate studies on the characteristics of infrared spectra of oil flame, an analytical study on flame spectra of various oil types and other combustible objects in outdoor space is carried out by establishing an all-flame infrared testing system with the spectral range of 1-14 μm. The results show that the signal of honeycomb briquette spectrum is the most special among all the spectra. After 5-layer decomposition of other fuel flame spectral signals by the db2 mother wavelet, low frequency components of 92# gasoline, 95# gasoline, 0# diesel oil, aviation kerosene and lube flame spectral features axe similar and there exist strong emission peaks in the vicinity of 1.2, 3.4 and 4.5 μm. The 5th layer detail coefficients of various fuel oil flame are in line with each other. The low frequency components and the detail coefficients of 92# gasoline and 0# diesel oil flame spectra have obvious features compared with those of other fuels （wood, alcohol and paper）. The experimental conclusion is of great significance in the detection and identification of oil flame based on spectral characteristic analysis.