等离子体氧化NO的试验研究及神经网络预测

Experimental Research and Neural Network Prediction of NO Oxidation Based on Plasma

NO的氧化对于提高湿式氨法的脱硝效率具有重要意义。通过试验研究考察了烟气流量、放电电流、温度、NO浓度和SO2浓度等参数对NO氧化效率的影响,结果表明,烟气流量、温度、NO浓度的增加和SO2的存在都会抑制NO的氧化,但SO2浓度的影响不大,而放电电流的增加会促进NO的氧化。基于试验研究,利用L-MBP神经网络建立了NO氧化率的预测模型,经过训练后对NO的氧化率进行了预测,预测结果与试验结果吻合度较好,说明L-MBP神经网络预测模型用于NO氧化效率的预测是可行的。

It is important to oxide NO to improve the efficiency of ammonia-based wet flue gas denitrification. The effects on NO oxidation efficiency were experimentally researched as to some parameters, such as flue gas flow, discharge current, temperature and concentration of NO and SO2. It was shown that the increase of flue gas flow, temperature and NO concentration and the presence of SO2 inhibited NO oxidation, but the impact of SO2 concentration was small, and the discharge current increase, on the contrary, did enhance the oxidation of NO. Based on the experimental study, the NO oxidation rate prediction model was built by using L-MBP neural network. The prediction results conformed very closely to the experimental results after training, which indicates that it is feasible to use the LMBP neural network prediction model to predict the efficiency of NO oxidation.