燃烧器燃烧高效低NOx排放建模与优化

摘要

燃烧器控制系统是实现燃烧器节能减排的核心,但由于燃烧器控制对象存在纯延迟、大滞后和多分布参数等特点,给控制过程造成困难.从燃烧器输入量到输出量的非线性关系入手,建立最小二乘支持向量机(LSSVM)燃烧效率和氮氧化物(NOx)排放量的数学模型,并结合遗传算法对反映燃烧状态的烟气O2量、CO量寻优,得到的O2量、CO量可作为燃烧器闭环控制的给定值,参与电子比调系统的优化控制.结果表明,利用LSSVM得到的预测结果的相对误差均小于0.8％,可实现较好燃烧状态预测;通过寻优的烟气O2量、CO量能够在保证燃烧效率略微降低的基础上,实现氮氧化物排放量的明显降低,使燃烧效率和氮氧化物排放量得到平衡,对燃烧器节能减排控制具有参考价值.%Combustor system is the core to achieve energy-saving and low environmental pollution,but it' s difficult to realize to control purpose for there are pure delay,large delay distribution parameters in combustor control object.Established LSSVM model of combustion efficiency and NOx emissions,combined GA to optimize the settings of O2 and CO in the electronic proportional closed-loop control system,which reflect the combustion state.The results showed that less than 0.8％ predicting performance were achieved with proper learning parameters choosing by PSO and optimal O2 and CO by GA process realize significantly lower NOx emissions while combustion efficiency was slightly decreased.It balanced combustion efficiency and NOx emissions,offer reference value in energy-saving and low environmental pollution of combustion control.