摘要
改性酚醛泡沫具有良好的吸附性能,可用于脱除气态污染物,需深入研究其脱除效率与各影响因素间关系。首次引入支持向量机建立改性酚醛泡沫吸附反应床模型,研究在不同温度、气质比、含氧量等影响因素下改性酚醛泡沫脱除二氧化硫、氮氧化物等气态污染物的脱除效果及最优反应条件。烟气模拟脱硫脱硝实验确定RBF-ε-SVM模型为反应器内最优污染物浓度分布预测模型,惩罚系数c=100,gamma因子g=0.1。基于最优模型的各影响因素实验表明:氧气含量在6%时改性酚醛泡沫脱硫脱硝效果最佳;随着质气比的增加脱除效果增强;反应温度在80℃内脱除SO_2和NO的效率随着烟气温度的升高而降低。该模型可用于改性酚醛泡沫吸附反应床的最优工况选择,反应器内浓度分布的在线监控,以及指导反应器的放大、中试。
A modified phenolic foam with excellent properties was used in this research to remove gaseous pollutants.The relationships between the removal rate of SO_2/NO and several influencing factors of such were studied using a support vector machine(SVM) that simulates the gaseous pollutants removal process by the modified phenolic foam.Processing experimental data on the removal of SO_2/NO by the modified phenolic foam with different mass gas ratios,oxygen contents,and temperaments were obtained.Taming the data to determine the RBF-ε-SVM model for optimal predictive models that punishing factor c=100 and gamma factor g=0.1.The RBF-ε-SVM model predicted the distribution of SO_2/NO concentration in the reactor.The predicting concentration was visualized through MATLAB software.The results showed that 6% oxygen in the flue was the most advantageous for removing SO_2/NO and a higher mass gas ratio was favorable for SO_2/NO removal.The removal efficiency of SO_2 and NO decreased as the gas temperature increased below 80 ℃.The model can be used to select the optimal operating conditions for the modified phenolic foam adsorption reaction bed,reactor concentration distribution prediction,and guidance for reactor scale-up.
出处
《环境工程学报》
CAS
CSCD
北大核心
2017年第5期2909-2914,共6页
Chinese Journal of Environmental Engineering
基金
国家自然科学基金资助项目(51178195)
关键词
改性酚醛泡沫
支持向量机
大气污染物控制
浓度预测
modified phenolic foam
support vector machine(SVM)
air pollutant control
concentration prediction