摘要
针对火电机组引风机故障频发的问题,提出了一种基于大数据分析的引风机故障预警方法。以某330 MW火电机组引风机为研究对象,提取了与其工作状态密切关联的11种特征信息,进行数据预处理后,结合BP神经网络建立了引风机状态预测模型,其误差满足工程要求。引入分布式控制系统(distributed control system,DCS)和安全仪表系统(safety instrumentation system,SIS),设计了一套引风机状态预警系统,并开发了此预警系统的可视化界面,当某一特征信息的实测值超过模型预测值的安全阈值后,系统会提醒运行人员进行检修和故障排查,实现对引风机故障的有效预警。
Aiming at the frequent occurrence of induced draft fan faults in thermal power units,a kind of induced draft fan fault warning method based on big data analysis was proposed.Taking the induced draft fan of a 330 MW thermal power unit as the research object,11 kinds of characteristic information closely related to its working state were extracted.After data pretreatment,a state prediction model of the induced draft fan was established based on BP neural network,and its error met the requirements of engineering.Using distributed control system(DCS)and safety instrumentation system(SIS),a set of induced draft fan state warning system was designed,and the visual interface of this warning system was developed.When the measured value of a certain feature information exceeds the safety threshold of the predicted value of the model,the system will remind the operator to carry out maintenance and troubleshooting,so as to realize an effective early warning of the induced draft fan fault.
作者
安吉振
郑福豪
刘一帆
陈衡
徐钢
AN Jizhen;ZHENG Fuhao;LIU Yifan;CHEN Heng;XU Gang(Beijing Key Laboratory of Pollutant Monitoring and Control in Thermoelectric Production,North China Electric Power University,Changping District,Beijing 102206,China)
出处
《发电技术》
CSCD
2023年第4期557-564,共8页
Power Generation Technology
基金
国家自然科学基金项目(51806062)
中央高校基本科研业务费(2020MS006)。
关键词
火力发电
火电机组
引风机
BP神经网络
故障预警
大数据
thermal power generation
thermal power unit
induced draft fan
BP neural network
fault warning
big data
