供热系统传感器数据过失误差侦破方法研究  

Research on data error detection method of heating system sensor

作　　者：李思璐 柳耀文 于景萍 傅强 刘海波 周志刚 薛普宁[4,5] 刘京 LI Silu;LIU Yaowen;YU Jingping;FU Qiang;LIU Haibo;ZHOU Zhigang;XUE Puning;LIU Jing(MCC Huatian Engineering and Technology Corporation,Nanjing 210019,China;Huadian Energy Thermal Network Testing and Technical Service Center,Harbin 150090,China;Heilongjiang Construction Engineering Group Company Limited,Harbin 150046,China;School of Architecture,Harbin Institute of Technology;Key Laboratory of Environment Science of Urban and Rural Residents in Cold Region and Technology Industry and Information Technology,Harbin Institute of Technology,Harbin 150090,China)

机构地区：[1]中冶华天工程技术有限公司,南京210019 [2]华电能源热网检测和技术服务中心,哈尔滨150090 [3]黑龙江省建工集团有限责任公司,哈尔滨150046 [4]哈尔滨工业大学建筑学院,哈尔滨150090 [5]哈尔滨工业大学寒地城乡人居环境科学与技术工业和信息化部重点实验室,哈尔滨150090

出　　处：《华电技术》2018年第11期1-4,77,共5页HUADIAN TECHNOLOGY

基　　金：黑龙江省自然科学基金重点项目(ZD2016010)

摘　　要：为提高集中供热系统上传数据的准确性,研究误差数据的识别方法,应用小波滤波、中值滤波、误差反向传播(BP)神经网络3种方法对供热系统换热站内的一级网流量传感器上传的数据进行过失误差侦破,并对比3种方法的侦破结果,得出中值滤波方法的过失误差侦破效果最好的结论。In order to improve the accuracy of data uploading by central heating system,the identification method of error data is studied. Three methods including wavelet filtering,median filtering and error back propagation( BP) neural network are applied by the heat exchange station in the heating system. The data uploaded by the first level network traffic sensor are used to detect the error,and the results detected by the three methods are compared. It is concluded that the error detection method with the median filtering is the best.

关 键 词：供热系统 小波滤波 中值滤波 误差反向传播神经网络 过失误差侦破 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] TU833.1[自动化与计算机技术—控制科学与工程]