论文部分内容阅读
火焰自由基对深入了解燃烧机理起着重要作用。通过数字成像研究火焰自由基光谱特征,并利用该特征参数建立极限学习机(ELM,Extreme Learning Machine)模型,以试验数据与数字仿真相结合的方法实现在生物质燃烧中对NOx的排放在线预测。用电子倍增CCD(EMCCD)相机采集火焰中的四种自由基OH*,CN*,CH*和C_2*的数字图像,采用模糊C均值聚类法(FCM,Fuzzy C-Means)进行图像分割并提取特征值,结合燃烧火焰温度,利用极限学习机进行NOx排放预测建模。采用燃气燃烧试验炉上的试验数据验证了预测模型的有效性。
Flame free radicals play an important role in understanding the combustion mechanism. The imaging characteristics of flame free radical spectrum were studied by digital imaging, and an ELM (Extreme Learning Machine) model was established to predict the emission of NOx in biomass combustion by a combination of experimental data and digital simulation . Digital images of the four free radicals OH *, CN *, CH * and C_2 * in the flame were acquired by using an electron-multiplying CCD (EMCCD) camera. The images were segmented using FCM (Fuzzy C-Means) The eigenvalues were extracted, and combined with the combustion flame temperature, the model of NOx emission prediction was established by using extreme learning machine. The test data on the gas burning test furnace is used to verify the effectiveness of the predictive model.