基于RBF神经网络的燃煤火电机组脱硝自动控制  

Automatic Denitration Control of Coal-fired Thermal Power Unit Based on RBF Neural Network

作　　者：何志瞧 李敏 梁龙飞 HE Zhiqiao;LI Min;LIANG Longfei(Zhejiang Zheneng Lanxi Power Generation Co.,Ltd.,Jinhua 321100,China;Beijing Huadian Tianchuang Intelligent Control Technology Co.,Ltd.,Beijing 102200,China)

机构地区：[1]浙江浙能兰溪发电有限责任公司,浙江金华321100 [2]北京华电天创智能控制技术有限公司,北京102200

出　　处：《机械与电子》2023年第11期17-21,共5页Machinery & Electronics

基　　金：浙江浙能兰溪发电有限责任公司科技项目(ZJTY10190943)。

摘　　要：燃煤火电机组脱硝过程中的数据样本较多,导致燃煤火电机组脱硝自动控制难度增加,因此提出基于RBF神经网络的燃煤火电机组脱硝自动控制方法。将燃煤火电机组脱硝入口烟气氮氧化物浓度、喷氨量、烟气入口温度和机组负荷作为数据样本训练RBF神经网络。通过敏感度法,分析输入数据样本值对其相应输出值的影响,优化RBF网络结构。将数据样本输入至优化后的RBF网络中,输出脱硝喷氨量计算结果,结合脱硝喷氨量完成燃煤火电机组脱硝自动控制。实验表明:火电机组负荷匀速变化状况下利用该方法可有效自动控制脱硝装置,降低出口氮氧化物浓度、喷氨量和氨气逃逸量,提升脱硝率;在不同负荷扰动情况下,该方法依旧能够有效控制脱硝装置,喷氨量和氨气逃逸量等均未超过设定值,证明该方法具有较优的脱硝自动控制效果。There are many data samples in the denitration process of coal-fired thermal power units,which makes it more difficult to control the denitration of coal-fired thermal power units automatically.Therefore,an automatic control method for the denitration of coal-fired thermal power units based on RBF neural network is proposed.The RBF neural network is trained by taking the concentration of NO_(x)in flue gas,ammonia injection amount,flue gas inlet temperature and unit load as data samples.By using sensitivity method,the influence of input data sample value on its corresponding output value is analyzed,and the RBF network structure is optimized.The data samples are input into the optimized RBF network,the calculation result of denitration injection ammonia is output,and the automatic denitration control of coal-fired thermal power unit is completed in combination with the denitration injection ammonia amount.Experimental results show that this method can effectively and automatically control the denitrification device under the condition of constant load change of thermal power unit,reduce the nitrogen oxide concentration at the outlet,the amount of ammonia injection and the amount of ammonia escape,and increase the denitrification rate.In the case of different load disturbances,the method can still effectively control the denitration device,and the amount of ammonia injection and the amount of ammonia escape do not exceed the set value,which proves that the method has a better automatic control effect of denitration.

关 键 词：RBF网络 燃煤火电机组 脱硝自动控制 氮氧化物 喷氨量 敏感度法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TM621[自动化与计算机技术—控制科学与工程]