基于主元分析的SCR烟气脱硝系统入口NOx生成量预测模型

摘要

Against such problems occurred in selective catalytic reduction (SCR)systems,such as large de-lay of ammonia inj ection control and not easy to realize automatic control,the NOx formation at entrance of the SCR denitration system was associated with variations of the unit main parameters.The principal com-ponent analysis method was applied to determine the main factors affecting the NOx formation.On this ba-sis,a dynamic model for NOx formation prediction at entrance of the SCR denitration system was estab-lished,using the nonlinear fitting method.The NOx formation prediction value was compared with the actu-al data.The results show that this model achieved a good prediction in the unit's steady state.While for un-steady state,there is a certain deviation between the model prediction value and the actual data,the predic-ted value was about 100 second ahead.Therefore,this established model can be used as the feedforward of ammonia spraying control for the SCR system to overcome the large delay and control the secondary air flow.%针对选择性催化还原(SCR)烟气脱硝系统喷氨量控制的大迟延特性，使得难以实现其自动控制的问题，将机组主要运行参数与 SCR 烟气脱硝系统入口 NOx 生成量相关联，通过主元分析确定影响N Ox 生成的主导因素.基于主元分析结果，通过非线性拟合方法建立了 N Ox 生成量预测模型，并将N Ox 生成量模型预测值与实测值进行对比.结果表明：机组处于稳态过程时，模型预测值与实测值吻合较好；机组处于动态过程时，模型预测值与实测值之间具有一定的偏差，模型预测值变化较实测值提前约100 s.因此，该预测模型可作为 SCR 烟气脱硝系统喷氨量控制前馈，从而克服系统的大迟延特性，同时可用于控制二次风挡板.