基于声学测温与人工神经网络的炉膛结渣在线监测方法  

The On-Line Monitoring of the Furance Slagging Based on Acoustic Pyrometry and Artificial Neural Network

作　　者：邓喆[1] 李庚生[1] 安连锁[1] 沈国清[1] 张世平[1] 

机构地区：[1]华北电力大学能源动力与机械工程学院,北京102206

出　　处：《现代电力》2011年第4期70-72,共3页Modern Electric Power

摘　　要：以某电厂锅炉炉膛结渣问题为研究背景,以实现炉膛受热面结渣监测为目的,提出了基于声学测温与人工神经网络的炉膛结渣在线监测方法。该监测方法结合非接触式测温工具——声学测温装置和BP神经网络,声学测温装置能够实时测量炉膛出口烟温,为炉膛结渣实时在线监测提供了前提条件,利用BP神经网络的非线性映射优点计算炉膛理想出口烟温,提出利用污染系数判断炉膛整体的结渣状况。在某电厂采集数据进行验证,结果表明:声学测温装置测量的炉膛理想出口烟温与BP神经网络预测的炉膛理想出口烟温相对误差在3%左右,能够满足工程要求。Through researching on a boiler furnace slagging,the on-line furnace heating surface slagging monitoring system based on acoustic pyrometry and artificial neural network is put forward.The method combines the non-contact temperature measuring tool,acoustic pyrometry device,with BP neural network.The acoustic pyrometry device can measure the furnace outlet gas temperature,which provides the precondition for the furnace slagging on-line monitoring.It can make use of BP neural network＇s nonlinear mapping calculation of the furnace ideal outlet gas temperature,and estimate slagging of the whole furnace with the pollution coefficient.To verify the method of slagging on-line calculation,the data of furnace ideal outlet gas temperatures from a power plant between on-line acoustic pyrometry and BP neural network prediction can meet the project requirement and the relative error is about 3%.

关 键 词：结渣 声学测温 BP神经网络 在线监测 炉膛出口烟温 

分 类 号：TM621.2[电气工程—电力系统及自动化]