数字孪生峡江水利枢纽水泵机组预测性维护关键技术研究及应用  

Research and application of key technologies for predictive maintenance of pump units in digital twin Xiajiang Hydro-junction

作　　者：张煜 傅韬 王颖 石强 徐曙光 毛鑫 ZHANG Yu;FU Tao;WANG Ying;SHI Qiang;XU Shuguang;MAO Xin(Jiangxi Provincial Xiajiang Hydro-junction Project Administration,Nanchang 330000,China;Jiangxi Provincial Water Conservancy Investment Jianghe Information Technology Co.,Ltd,Nanchang 330000,China)

机构地区：[1]江西省峡江水利枢纽工程管理局,江西南昌330000 [2]江西省水投江河信息技术有限公司,江西南昌330000

出　　处：《水利信息化》2024年第6期14-22,共9页Water Resources Informatization

摘　　要：为探索实现大型泵站水泵机组安全监测与预测性维护,保障机组安全稳定高效运行,基于数字孪生峡江水利枢纽工程先行先试项目,针对江西省峡江水利枢纽工程库区同江电排站水泵机组安全监测手段不足、设备故障无法预知、故障定位难等问题,研究大型泵站机组预测性维护关键技术与具体实现。依托过程机器多能域耦合动力学建模、张量自适应分解信号处理、人工智能训练等关键技术,在峡江水利枢纽数字孪生平台上实现设备状态在线监测、设备故障诊断实时预报、设备健康状态自动化预警、设备监测数据及预报预警成果可视化展示等功能。形成的技术体系与建设成果,适用于水利工程转动机电设备安全运维,为水利工程安全、调度运行、维修管护提供决策支撑。To ensure the safe,stable,and efficient operation of large-scale pump units in pump stations,this study explores predictive maintenance techniques through the pioneering application of a digital twin framework for the Xiajiang Hydro-junction.Addressing the limitations of safety monitoring,inability to predict faults,and difficulty in fault localization in Tongjiang Electric Drainage Station pump units within the Xiajiang Hydro-junction reservoir area,the research focuses on key technologies and their implementation for predictive maintenance of large pump station units.Relying on multi-domain coupling dynamics modeling,tensor adaptive decomposition signal processing,and artificial intelligence training,the digital twin platform of the Xiajiang Hydro-junction enables real-time monitoring of equipment status,fault diagnosis and forecasting,automated health status warnings,and visualization of monitoring data and results.The developed technical system and outcomes are applicable to the safe operation and maintenance of rotating electromechanical equipment in hydraulic projects,providing decision-making support for the safety,scheduling,and maintenance of hydraulic engineering.

关 键 词：水泵机组 预测性维护 数字孪生 多能域耦合 信号处理 人工智能 

分 类 号：TV675[水利工程—水利水电工程] TP391.98[自动化与计算机技术—计算机应用技术]