标题：Data driven induction motor condition identification and fault prediction
作者：Guo, Lu ;Yu, Zhongqing ;Yu, Jianqi ;Liu, Chuang
作者机构：[Guo, Lu ;Yu, Zhongqing ;Liu, Chuang ] School of Data Science and Software Engineering, Qingdao University, Qingdao, Shandong, China;[Yu, Jianqi ] Col 更多
会议名称：2nd International Conference on Big Data Technologies, ICBDT 2019
会议日期：August 28, 2019 - August 30, 2019
来源：ACM International Conference Proceeding Series
摘要：With the development of technologies such as sensing and the scale of industrial production, the equipment data can be gathered massively. For induction motors, the health data can be collected is sufficient, but the fault data is not easy to obtain. Therefore, the focus of this paper was determined to identify motor operating conditions and predict possible faults based on motor health data. In this case, an induction motor condition model consisting of a state recognizer and adaptive thresholds was proposed. The health data was used for the training of the induction motor condition model, and an improved SOM-FCM Two-Layer clustering method was used to solve the problem of obtaining the motor data without label. Finally, the validity of the model and method was verified by normal motor variable load state identification and rotor broken motor fault prediction, and the accuracy of 97.5% and 90.2% was obtained respectively.
© 2019 Association for Computing Machinery.