摘要
飞灰含碳量的准确测量是提高锅炉燃烧效率的基础,针对目前飞灰含碳量测量装置速度、精度不理想的情况,提出了基于互信息变量选取的神经网络飞灰含碳量预测模型,并通过PLC和上位机组合的方式对神经网络进行在线监控。首先,介绍了锅炉燃烧机理,针对影响飞灰含碳量的因素进行分析,对机理分析得到的影响因素通过互信息进行选取,得到飞灰含碳量软测量模型建立所需要的辅助变量;然后针对选取得到的辅助变量进行数据预处理,包含数据去重、数据滤波、去异常值等,以处理之后的数据为输入建立神经网络模型;最后,通过PLC的SCL语言对建立的神经网络模型进行编程实现,并通过上位机组态软件WinCC进行飞灰含碳量进行在线监控。结果表明本文所建立的动态模型相较于传统的飞灰含碳量静态神经网络模型具有更高的实用性和准确性,可对现场采集的数据进行实时计算得到飞灰含碳量值并进行校正;PLC测量装置有着良好的预测精度与较高的预测速度,能够用于现场飞灰含碳量测量。
Accurate measurement of carbon content in fly ash is the basis for improving boiler combustion efficiency.In view of the current unsatisfactory speed and accuracy of fly ash carbon content measuring device,a prediction model of carbon network fly ash carbon content based on mutual information variable was proposed,and online monitoring of the neural network was achieved through the combination of PLC and host computer.Firstly,the boiler combustion mechanism is introduced,and the factors affecting the carbon content of fly ash are analyzed.The influencing factors obtained by the mechanism analysis are selected by mutual information,and the auxiliary variables needed for the soft measurement model of fly ash carbon content are obtained.The obtained auxiliary variables are selected for data preprocessing,including data de-duplication,data filtering,and outliers,and the neural network model is established by inputting the processed data.Finally,the established neural network model is implemented by the SCL language of PLC,and the carbon content of the fly ash is monitored online by the host computer configuration software WinCC.The results show that the dynamic model established in this paper has higher practicability and accuracy than the traditional fly ash carbon-free static neural network model.The data collected in the field can be calculated in real time to obtain the carbon content of fly ash and carry out correction;PLC measuring device has good prediction accuracy and high prediction speed,and can be used for on-site fly ash carbon content measurement.The calculation speed and prediction accuracy of this device are better than those of previous measurement devices.
作者
乔源
王建峰
杨永存
赵文杰
QIAO Yuan;WANG Jianfeng;YANG Yongcun;ZHAO Wenjie(School of Control and Computer Engineering,North China Electric Power University,Baoding 071003,China;Shanxi Zhangshan Power Generation Co.,Ltd.,Changzhi 046021,China;National Electric Power Investment Group Ningxia Energy Aluminum Industry Linhe Power Generation Branch,Lingwu 751400,China)
出处
《电力科学与工程》
2019年第11期55-61,共7页
Electric Power Science and Engineering
关键词
飞灰含碳量
神经网络
互信息
PLC通讯
WINCC组态
fly ash carbon content
neural network
mutual information
PLC communication
WinCC configuration