摘要
为提前识别动车组运行过程中的故障隐患,文章基于故障预测及健康管理理论,通过整合动车组不同场景的多源异构数据,构建“车-地”一体化大数据平台;并将设备故障机理与人工智能算法相结合,构建一种动车组关键部件故障预警预测模型,以部件关键物理特性来反映其工作状态并提前识别潜在故障。通过大数据平台及牵引电机的故障预警及温度预测模型的应用,CR400AF型复兴号动车组牵引电机故障率显著下降,由平均每百万公里0.5件降至约平均每百万公里0.1件。
In order to identify potential fault hazards in the operation of EMU in advance,based on the theory of prognostics health management,this paper constructs a vehicle-ground integrated big data platform by integrating multi-source heterogeneous data of different scenarios of the EMU.It uses a method of combining equipment fault mechanism and artificial intelligence algorithm to construct a fault early warning and prediction model of key components of EMU,which reflects the working status of a component with its key physical properties and identify potential faults in advance.Through the application of the big data platform and the traction motor fault early warning and temperature prediction model,traction motor fault rate of CR400AF Fuxing EMU decreased significantly from 0.5 pieces/10^(6) km to about 0.1 pieces/10^(6) km.
作者
吴臻易
WU Zhenyi(CRRC Qingdao Sifang Co.,Ltd.,Qingdao,Shandong 266111,China)
出处
《控制与信息技术》
2021年第5期91-96,共6页
CONTROL AND INFORMATION TECHNOLOGY
基金
山东省重点研发计划(国际科技合作)(2019GHZ004)。
