专家知识嵌入的高泛化可解释高速列车转向架轴箱轴承故障诊断大模型

doi:10.3901/JME.260720

摘要/Abstract

摘要： 轴箱轴承是高速列车走行部核心部件，其高载荷、高转速特性导致故障易引发安全风险。现有基于深度学习的故障诊断模型存在两大局限：泛化能力不足，难以适配高速列车多变运行工况，并且诊断结果缺乏可解释性，无法满足智能运维的可信度需求。受大模型在语义理解与任务泛化上的优势启发，提出专家知识嵌入的高泛化可解释故障诊断大模型，通过监测数据、任务文本、专家知识多模态融合与集成推理，实现诊断结果和依据文本双输出。核心优势在于：① 以检索增强生成技术实现专家知识高效嵌入，弥补数据驱动模型的知识缺口；② 通过多模态特征对齐，提升跨场景泛化能力；③ 集成多模型诊断依据，依托推理大模型完成诊断归因。试验结果表明：所提模型在跨工况泛化任务中平均诊断准确率达89.8%，跨数据泛化任务中达87.1%，均显著优于传统域适应/域泛化模型，同时能输出符合专业逻辑的诊断依据，有效提升智能运维的可靠性与可解释性。

关键词: 故障诊断, 专家知识, 健康管理, 智能运维, 大模型

Abstract: Axlebox bearings constitute a core component of high-speed train running gear. The high-load, high-speed characteristics render failures prone to triggering safety risks. Existing deep learning-based fault diagnosis models exhibit two major limitations：insufficient generalization capability, making them difficult to adapt to the variable operating conditions of high-speed trains; and a lack of interpretability in diagnostic results, failing to meet the credibility requirements of intelligent maintenance operations. Inspired by large models’ strengths in semantic understanding and task generalization, a highly generalisable, explainable fault diagnosis large model incorporating expert knowledge is proposed. This model achieves dual outputs of diagnostic results and supporting rationale through multimodal fusion and integrated reasoning of monitoring data, task texts, and expert knowledge. Its core strengths lie in：① efficiently embedding expert knowledge via retrieval-enhanced generation to bridge the knowledge gap in data-driven models; ② enhancing cross-scenario generalization through multimodal feature alignment; ③ integrating multi-model diagnostic rationale to perform diagnostic attribution via a reasoning large model. Experimental results demonstrate that the proposed model achieves an average diagnostic accuracy of 89.8% in cross-operating-condition generalization tasks and 87.1% in cross-data generalization tasks. Both figures significantly outperform traditional domain adaptation models, which generates diagnostics aligned with professional logic, effectively enhancing the reliability and explainability of intelligent operations and maintenance.

Key words: fault diagnosis, expert knowledge, prognostics and health management, intelligent maintenance, large model

中图分类号:

TG156

郭志斌, 谢劲松, 王田天, 潘彤阳, 牛步钊. 专家知识嵌入的高泛化可解释高速列车转向架轴箱轴承故障诊断大模型[J]. 机械工程学报, 2026, 62(16): 355-368.

GUO Zhibin, XIE Jingsong, WANG Tiantian, PAN Tongyang, NIU Buzhao. Large Models with Embedded Expert Knowledge for Highly Generalized Interpretable High-speed Train Bogie Axle-box Bearing Fault Diagnosis[J]. Journal of Mechanical Engineering, 2026, 62(16): 355-368.

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: http://www.cjmenet.com.cn/CN/10.3901/JME.260720

http://www.cjmenet.com.cn/CN/Y2026/V62/I16/355

参考文献

[1] 中国政府网. 4.8万公里！我国高铁运营里程再创新纪录[EB/OL]. (2025-01-02)[2025-06-06]. https://www.gov.cn/yaowen/liebiao/202501/content_6995868.htm. Chinese Government Website. 48000 kilometers! our country’s high-speed rail operating mileage has set a new record[EB/OL]. (2025-01-02)[2025-06-06]. https://www.gov.cn/yaowen/liebiao/202501/content_6995868.htm.
[2] 杨绍普，顾晓辉，刘永强，等. 转向架关键运动部件动力学机理与故障诊断研究综述[J]. 机械工程学报，2023，59(20)：225-243. YANG Shaopu，GU Xiaohui，LIU Yongqiang，et al. Review of dynamic mechanism and fault diagnosis for key rotating components of bogies[J]. Journal of Mechanical Engineering，2023，59(20)：225-243.
[3] 徐磊，王田天，谢劲松，等. 高速列车轴箱轴承故障特征与诊断技术研究[J]. 中南大学学报(自然科学版)，2022，53(5)：1600-1613. XU Lei，WANG Tiantian，XIE Jingsong，et al. Research on fault characteristics and diagnosis technology of axle-box bearing of high-speed train[J]. Journal of Central South University (Science and Technology)，2022，53(5)：1600-1613.
[4] 刘永强，王宝森，杨绍普. 含外圈故障的高速列车轴承转子系统非线性动力学行为分析[J]. 机械工程学报，2018，54(8)：17-25. LIU Yongqiang，WANG Baosen，YANG Shaopu，et al. Nonlinear dynamic behaviors analysis of the bearing rotor system with outer ring faults in the high-speed train[J]. Journal of Mechanical Engineering，2018，54(8)：17-25.
[5] 李彦夫，韩特. 基于深度学习的工业装备PHM研究综述[J]. 振动、测试与诊断，2022，42(5)：835-847，1029. LI Yanfu，HAN Te. Deep learning based industrial equipment prognostics and health management：A review[J]. Journal of Vibration，Measurement & Diagnosis，2022，42(5)：835-847，1029.
[6] 雷亚国，贾峰，孔德同，等. 大数据下机械智能故障诊断的机遇与挑战[J]. 机械工程学报，2018，54(5)：94-104. LEI Yaguo，JIA Feng，KONG Detong，et al. Opportunities and challenges of machinery intelligent fault diagnosis in big data era[J]. Journal of Mechanical Engineering，2018，54(5)：94-104.
[7] 罗宏林，柏林，侯东明，等. 有限变工况特征迁移学习方法及其在高速列车轴箱轴承故障诊断中的应用[J]. 仪器仪表学报，2022，43(3)：132-145. LUO Honglin，BO Lin，HOU Dongming，et al. A transfer learning method for bearing diagnosis under finite variable working conditions and its application in train axle-box bearings fault diagnosis[J]. Chinese Journal of Scientific Instrument，2022，43(3)：132-145.
[8] TANG Baoping，HUANG Qingqing，DENG Lei，et al. Research progess and challenges of wireless sensor networks for machinery equipment condition monitoring[J]. Journal of Vibration，Measurement & Diagnosis，2014，34(1)：1-7.
[9] ZHAO Chao，ZIO E，SHEN Weiming，et al. Domain generalization for cross-domain fault diagnosis：An application-oriented perspective and a benchmark study[J]. Reliability Engineering & System Safety，2024，245：109964.
[10] 雷亚国，李熹伟，李响，等. 面向机械设备通用健康管理的智能运维大模型[J]. 机械工程学报，2025，61(6)：1-13. LEI Yaguo，LI Xiwei，LI Xiang，et al. Research on large model for general prognostics and health management of machinery[J]. Journal of Mechanical Engineering，2025，61(6)：1-13.
[11] 雷亚国，杨彬，杜兆钧，等. 大数据下机械装备故障的深度迁移诊断方法[J]. 机械工程学报，2019，55(7)：1-8. LEI Yaguo，YANG Bin，DU Zhaojun，et al. Transfer diagnosis method for machinery in big data era[J]. Journal of Mechanical Engineering，2019，55(7)：1-8.
[12] 邓飞跃，董少飞，顾晓辉. 基于ITR-Net多源域迁移学习的高铁轴箱轴承故障诊断研究[J]. 工程科学与技术，2026，58(1)：324-333. DENG Feiyue，DONG Shaofei，GU Xiaohui. A study on fault diagnosis of high-speed railway axle-box bearing based on ITR-Net multi-source domain migration learning[J]. Advanced Engineering Sciences，2026，58(1)：324-333.
[13] YANG Buyao，WANG Tiantian，XIE Jingsong，et al. Deep adversarial hybrid domain-adaptation network for varying working condition fault diagnosis of high-speed train bogie[J]. IEEE Transactions on Instrumentation and Measurement，2023，72：3517510.
[14] ZHANG Tinaci，CHEN Jinglong，XIE Jingsong，et al. SASLN：Signals augmented self-taught learning networks for mechanical fault diagnosis under small sample condition[J]. IEEE Transactions on Instrumentation and Measurement，2021，70：3506511.
[15] QIN Yi，LIU Hongyu，MAO Yongfang，et al. Faulty rolling bearing digital twin model and its application in fault diagnosis with imbalanced samples[J]. Advanced Engineering Informatics，2024，61：102513.
[16] 邵海东，颜深，刘政武，等. 面向小样本场景的风机多增量故障诊断[J]. 机械工程学报，2025，61(10)：230-240. SHAO Haidong，YAN Shen，LIU Zhengwu，et al. Multi-incremental fault diagnosis of wind turbine towards small sample scenario[J]. Journal of Mechanical Engineering，2025，61(10)：230-240.
[17] 陈晓红，傅文润，刘朝明，等. 人工智能大模型在电力设备运维场景中的应用探讨[J]. 中国工程科学，2025，27(1)：180-192. CHEN Xiaohong，FU Wenrun，LIU Chaoming，et al. Application of artificial intelligence large language model in power equipment operation and maintenance[J]. Strategic Study of CAE，2025，27(1)：180-192.
[18] WANG Jie，LI Tianrui，YANG Yan，et al. DiagLLM：Multimodal reasoning with large language model for explainable bearing fault diagnosis[J]. Science China Information Sciences，2025，68：160103.
[19] TAO Laifa，LI Shuai，NING Guoao，et al. LLM-based framework for bearing fault diagnosis[J] Mechanical Systems and Signal Processing，2025，224：112127.
[20] MA Yunfei，ZHENG Shuai，YANG Zheng，et al. A knowledge-graph enhanced large language model-based fault diagnostic reasoning and maintenance decision support pipeline towards industry 5.0[J]. International Journal of Production Research，2025，5(3)：1-22.
[21] TAO Laifa，LI Shangyu，LIU Haifei，et al. An outline of prognostics and health management large model：Concepts，paradigms，and challenges[J]. Mechanical Systems and Signal Processing，2025，232(1)：112683.
[22] BECHHOEFER E. Condition based maintenance fault database for testing diagnostics and prognostic algorithms (2013) [EB/OL]. https://www.mfpt.org/fault-data-sets/.
[23] SMITH W A，RANDALL R B，et al. Rolling element bearing diagnostics using the case western reserve university data：A benchmark study[J]. Mechanical Systems and Signal Processing，2015，64-65：100-131.
[24] LESSMEIER C，KIMOTHO J K，ZIMMER D，et al. Condition monitoring of bearing damage in electromechanical drive systems by using motor current signals of electric motors：A benchmark data set for data-driven classification[C]// European Conference of the PHM Society 2016，Bilbao，Spain. PHM Society， 2016.
[25] QIN Yong，WANG Yiran，WANG Biao，et al. An in-depth tutorial on BJTU-RAO bogie datasets for fault diagnosis[J]. IEEE Access，2025，13：60879-60888.
[26] ZHANG Tianci，CHEN Jinglong，HE Shuilong，et al. Prior knowledge augmented self-supervised feature learning for few-shot intelligent fault diagnosis of machines[J]. IEEE Transactions on Industrial Electronics，2022，69：10573-10584.
[27] CHEN Zhihao，CEN Jian，XIONG Jianbin，et al. Rolling bearing fault diagnosis using time-frequency analysis and deep transfer convolutional neural network[J]. IEEE Access，2020，8：150248-150261.
[28] LI Haoliang，PAN S，WANG Shiqi，et al. Domain generalization with adversarial feature learning[C]// IEEE/CVF Conference on Computer Vision and Pattern Recognition，Salt Lake City，USA. 2018.
[29] WANG Zhiyong，KANG D K. P-norm attention deep coral：Extending correlation alignment using attention and the p-norm loss function[J]. Applied Sciences，2021，11(11)：5267.
[30] HAN Te，LI Yanfu，QIAN Min，et al. A hybrid generalization network for intelligent fault diagnosis of rotating machinery under unseen working conditions[J]. IEEE Transactions on Instrumentation and Measurement，2021，70：1-11.
[31] RAGAB M，CHEN Zhenghua，ZHANG Wenyu，et al. Conditional contrastive domain generalization for fault diagnosis[J]. IEEE Transactions on Instrumentation and Measurement，2022，71：1-12.
[32] ZHAO Chao，SHEN Weiming. A domain generalization network combing invariance and specificity towards real-time intelligent fault diagnosis[J]. Mechanical Systems and Signal Processing，2022，173：108990.
[33] 邵海东，肖一鸣，钟翔，等. 面向可信机械故障诊断的模型校准方法[J]. 机械工程学报，2025，61(17)：114-123. SHAO Haidong，XIAO Yiming，ZHONG Xiang，et al. Model calibration method for trustworthy mechanical fault diagnosis[J]. Journal of Mechanical Engineering，2025，61(17)：114-123.