摘要
为实现泵站设备的智能化管理和预测性维护,文章以博斯腾湖西泵站为研究对象,构建了基于数字孪生技术的设备管理平台。采用“三层耦合、双向映射”的创新架构,开发了设备故障诊断预警模型、关键部件趋势预测模型和主机组安全评价模型;通过LSTM-Attention双重神经网络与迁移学习相结合的方法,提高了故障诊断的准确性;基于改进的威布尔分布模型,实现了水导轴承寿命的动态预测。实践表明:该数字孪生模型显著提升了泵站设备管理水平,为大型泵站智能化运维提供了新思路。
In order to realize the intelligent management and predictive maintenance of pumping station equipment,this study takes Bosten Lake West Pumping Station as the research object,and builds an equipment management platform based on digital twin technology.Using the innovative architecture of “three-layer coupling and bidirectional mapping”,the equipment fault diagnosis and early warning model,key component trend prediction model and host group safety evaluation model are developed.By combining LSTM-Attention dual neural network with transfer learning,the accuracy of fault diagnosis is improved.Based on the improved Weibull distribution model,the life of water guide bearing is predicted dynamically.The practice shows that the digital twin model significantly improves the equipment management level of pumping station and provides a new idea for intelligent operation and maintenance of large pumping station.
作者
姬疆燕
黄哲
罗智敏
洪亚东
刘震
JI Jiangyan;HUANG Zhe;LUO Zhimin;HONG Yadong;LIU Zhen(Kongqi River Water Conservancy Management Center,Kaidu Tarim River Basin,Korla 841000,China;South-to-North Water Div(Jiangsu)ShuzhiTechnology Co.,Ltd.,Nanjing 210019,China)
关键词
博斯腾湖
西泵站
数字孪生技术
模型构建
应用效果
Bosten Lake
west pumping station
digital twin technology
model construction
application effect
