摘要
为克服现有动车组故障预测与健康管理(PHM)在多源数据融合、动态工况适应性、模型泛化能力等方面的不足,基于数字孪生技术,提出适用于动车组的PHM系统建模与应用框架。该框架以物理层、数据层、模型与功能层、服务层为核心,结合多体动力学仿真、实时数据同化、机器学习算法,实现对动车组关键系统(如牵引系统、转向架、空调系统等)的健康状态评估与精准故障预测。研究结果表明,构建的数字孪生PHM系统可有效提升故障预测的准确性,优化维护决策,通过视情维修策略,降低运维成本和维护作业负担。实际案例分析结果显示,该方法可在高速铁路复杂运行环境中实现更智能化、精细化的动车组运维管理,为高速铁路车辆全生命周期的安全可靠运行提供坚实的技术支撑。
To address the shortcomings of existing EMU Prognostics and Health Management(PHM)methods in terms of multi-source data fusion,dynamic operating condition adaptation,and model generalization,this paper proposes a digital-twin-based PHM modeling and application framework specifically designed for EMU trains.The framework comprises four layers,namely physical,data,model&function,and service layers.It integrates multi-body dynamics simulation,real-time data assimilation,and machine learning algorithms to assess the health status and accurately predict faults in key subsystems(e.g.traction,bogies and air conditioning).The results demonstrate that the digital twin PHM system significantly improves fault prediction accuracy and optimizes maintenance decisions,thus reducing both maintenance costs and labor intensity through conditionbased maintenance strategies.Furthermore,a real-world case study reveals that the proposed method enables a more intelligent and refined operation and maintenance strategy under complex high-speed railway operating environments,providing robust technical support for the safe and reliable operation of high-speed trains throughout their life cycles.
作者
吴奇
贾志凯
朱晓敏
甄鹏
李燕
崔言杰
WU Qi;JIA Zhikai;ZHU Xiaomin;ZHEN Peng;LI Yan;CUI Yanjie(School of Mechanical,Electronic and Control Engineering,Beijing Jiaotong University,Beijing 100044,China;Institute of Computing Technology,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China;Postgraduate Department,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China;System Software Product Department,Inspur Electronic Information Industry Co.,Ltd.,Ji’nan Shandong 250014,China)
出处
《中国铁路》
北大核心
2026年第1期43-50,共8页
China Railway
基金
国家自然科学基金资助项目(U2268205)
中国铁道科学研究院集团有限公司科研开发基金项目(2023YJ130)。
关键词
数字孪生
动车组
PHM
智能运维
建模方法
digital twin
EMU
PHM
intelligent operation and maintenance
modeling method
