基于运行数据的锅炉故障自动检测技术研究  

Research on Boiler Fault Automatic Detection Technology Based on Operation Data

作　　者：郭永谦 GUO Yongqian(Qinghai Special Equipment Inspection Institute,Xining 810000,China)

机构地区：[1]青海省特种设备检验所,青海西宁810000

出　　处：《机械制造与自动化》2022年第4期237-240,共4页Machine Building & Automation

摘　　要：为实时检测锅炉的运行数据,提高其连续工作性能,提出一种基于运行数据的锅炉故障自动检测方法。通过时间序列方式收集锅炉稳态数据,并归一化处理;根据处理后数据属性的均值、方差得出数据的集合与方差矩阵,提取出其数据特征;通过标准化方式得出特征数据各点之间的距离与距离矩阵,采用欧式距离与绝对距离测量融合方式检测出分布密集数据的故障数据,并利用距离度量函数,完成分布稀疏数据的故障检测,从而完成锅炉整个故障数据自动检测。实验结果表明:该方法锅炉故障检测结果精度高,故障检测消耗时间极短,具备良好鲁棒性。In order to detect the boiler operation data in real time and improve the continuous working performance,an automatic detection method of boiler fault is proposed based on operation data.The steady-state data of boiler are collected by time series and normalized.According to the mean and variance of the processed data attributes,the data set and variance matrix are obtained,and the data features are extracted.The distance and distance matrix between each point of feature data are gained through standardization.The fault data of distributed dense data is detected by fusion of euclidean distance and absolute distance measurement.The fault detection of distributed sparse data is completed by using distance measurement function.The experimental results show that the proposed method has high accuracy,short fault detection time and good robustness.

关 键 词：锅炉 数据采集 数据特征提取 故障数据 自动检测 数据点 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]