磨齿机机械手电力变压器运行状态检测研究  

Research on Testing of Running Status of Power Transformer of Gear Grinder Manipulator

作　　者：张文娅 张鹏[2] 高祥斌[3] ZHANG Wen-ya;ZHANG Peng;GAO Xiang-bin(School of Automation,Beijing Information Science and Technology University,Beijing 100083,China;Shandong Jiaotong University,Shandong Jinan 250357,China;Linyi University,Shandong Linyi 273400,China)

机构地区：[1]北京信息科技大学自动化学院,北京100083 [2]山东交通学院,山东济南250357 [3]临沂大学,山东临沂273400

出　　处：《机械设计与制造》2020年第12期252-255,共4页Machinery Design & Manufacture

基　　金：山东省软科学课题—原生态原则下山东省“互联网+海绵城市”智慧化建设研究立项编号(2018RKB01381)。

摘　　要：磨齿机机械手电力变压器设备故障状态众多,需要有监督约束分类条件,但约束条件过多导致关联性无法确定。采用时间序列与滑动窗口对磨齿机自动化装料机械手电力变压器在线运行状态监测数据进行过滤,标记异常状态的发生时间及类型,构建候选异常状态参量集合的甄别模型;运用无监督聚类方法简化异常状态参量之间的关联关系,避免参量间关联性无法确定的问题,运用滑动窗口和聚类算法线检测,使之适用于设备运行状态在线监测数据流,实现在线监测数据流中异常的快速检出。实验结果表明:这里方法能够在线检测出磨齿机机械手电力变压器设备所存在的隐患,相较于传统的人工检测和停电试验的方法,更加实用、有效。There are many failure states of the gear grinder manipulator equipment,which requires supervised and constrained classification conditions,but too many constraints make the correlation undetermined.The time series and sliding window are used to filter the online monitoring status data of the power transformer of the automatic charging manipulator of the gear grinding machine,mark the occurrence time and type of the abnormal state,and construct the screening model of the candidate abnormal state parameter set;use the unsupervised clustering method Simplify the relationship between abnormal state parameters and avoid the problem that the correlation between parameters cannot be determined.Use sliding window and clustering algorithm line detection to make it suitable for online monitoring data flow of equipment operating status and realize abnormal monitoring of online data flow.Quickly check out.The experimental results show that the method can detect the hidden dangers of the power transformer equipment of the gear grinding machine manipulator,which is more practical and effective than the traditional manual detection and power failure test methods.

关 键 词：在线检测 有监督 时间序列 滑动窗口 k-means聚类方法 关联性 

分 类 号：TH16[机械工程—机械制造及自动化] TP391.41[自动化与计算机技术—计算机应用技术]