基于支持向量机的选择性催化还原脱硝系统运行优化  

Operation optimization of selective catalytic reduction system based on support vector machine

作　　者：王桂林[1] 张宇[1] 王森[1] 边疆[1] 薛泽海[1] 孙国通 

机构地区：[1]国网天津市电力公司电力科学研究院,中国天津300010 [2]天津市电力科技发展公司,中国天津300022

出　　处：《环境工程学报》2015年第10期5011-5016,共6页Chinese Journal of Environmental Engineering

摘　　要：建立了锅炉运行过程NOx生成和脱硝效率的预测模型以及脱硝成本的优化模型。计算分析了选择性催化还原（SCR）脱硝系统入口NOx量、机组负荷、锅炉排烟氧量、喷氨量、一、二次风量及配比、煤质等因素对脱硝运行的影响规律,确定了不同运行条件下锅炉脱硝运行成本。利用遗传算法,以脱硝系统总运行成本最小为目标,对喷氨量进行优化,得到最佳喷氨量。结果表明,该电厂实际喷氨量比预测出的最佳喷氨量平均低大约1%~3.5%左右。In this study,models for NOxemission prediction and denitration costs calculation were established using support vector machine. The influences of the factors including selective catalytic reduction system inlet NOxcontent,unit load,oxygen content of boiler exhaust gas,ammonia injection rate,air conditioning,coal type on NOxreduction were analyzed. The denitration costs under different running conditions were calculated.The optimized ammonia injection rates were determined using genetic algorithm. Results show that the actual injection rate of ammonia is 1% —3. 5% lower than the optimal value.

关 键 词：选择性催化还原脱硝系统 喷氨量 NOX排放 脱硝效率 成本 

分 类 号：X511[环境科学与工程—环境工程]