基于代数迭代算法的燃烧火焰温度场和气体浓度场重建研究

基于TDLAS(tunable diode laser absorption spectroscopy)技术，以水汽作为目标气体，采用直接吸收的测量方式，探测了甲烷空气预混平焰炉燃烧区域水汽的吸收光谱信号，通过ART(algebraic reconstruction technique)代数迭代算法对燃烧场温度和水汽浓度分布进行了模拟重建和实验研究，模拟重建采取5×5共25个网格的正方形重建区域，假定25个网格的一个温度浓度二维分布，模拟28条激光束从不同的角度方位穿越重建区域，得到模拟射线下的投影值，经ART算法重建，结果显示温度场和水汽浓度场的重建偏差均在1%以内。实验采用分布反馈式激光器作为光源，选取H₂O的7 153.722，7 153.748和7 154.354 cm-1三条吸收线作为测温谱线，其中前两条线不区分作为一条吸收线来处理。使用平移台多方位平行扫描，共获取30路光谱吸收信号，经数据处理、算法重建和双线比值法测温原理得到了圆形平焰炉16个不同区域的温度浓度值，且炉面偏向中心区域温度浓度值较高，边缘较小，结果表明代数迭代算法能够很好地实现燃烧区域温度场和水汽浓度场的反演。

Reconstruction Research for Gas Concentration and Temperature of Flame Based on Algebraic Reconstruction Technique

We specify water vapor among combustion products as the target gas based on tunable diode absorption spectroscopy in this paper. The direct absorption signals of water vapor after being processed can be used to calculate the gas concentration distributions and temperature distributions of the combustion region of methane and air flat flame furnace via algebraic reconstruction technique (ART). In the numerical simulation, reconstruction region is a grid of five by five, we assume a temperature and water vapor concentration distribution of 25 grid, then simulate different direction laser rays which cross the combustion region, generating projection of each ray, by ART reconstruction algorithm, it turns out that the temperature and water vapor distribution reconstruction error is less than 1%. In the experiment, we chose a distributed-feedback laser to scan the target gas H₂O 7 153.722，7 153.748 and 7 154.354 cm-1 as absorbtion line pair to measure temperature of the flame，we consider the former two line as one absorbtion line. By Stages multi-directional scanning, the authors abtain 16 different regions distributin of temperature and gas concentration of furnace when we collecte 30 different angle data by spectral data processing, reconstruction algorithm, two absorbtion line ratio method for temperature sensing, finding the temperature and water concentration are higher in the center than in the edge, it turns out that the reconstruction algorithm is good enough to achieve the distributions of gas concentration and temperature of the combustion region.