Bearing life prediction method based on dynamic knowledge embedding

doi:10.11772/j.issn.1001-9081.2025070866

Journal of Computer Applications ›› 2026, Vol. 46 ›› Issue (7): 2347-2354.DOI: 10.11772/j.issn.1001-9081.2025070866

• Frontier and comprehensive applications • Previous Articles

Bearing life prediction method based on dynamic knowledge embedding

Jing LIU¹^,², Fengfeng LYU², Wei NIU³, Haipeng JI¹^,⁴(), Jian WU⁵, Xiao ZHANG⁵

^1.State Key Laboratory of High Performance Roll Materials and Composite Forming （Tianjin University），Tianjin 300072，China
^2.School of Artificial Intelligence，Hebei University of Technology，Tianjin 300401，China
^3.CITIC Dicastal Company Limited，Qinhuangdao Hebei 066011，China
^4.School of Materials Science and Engineering，Hebei University of Technology，Tianjin 300401，China
^5.Tianjin Research Institute of Electric Science Company Limited，Tianjin 300180，China

Received:2025-08-01 Revised:2025-09-08 Accepted:2025-09-09 Online:2025-11-05 Published:2026-07-10
Contact: Haipeng JI
About author:LIU Jing， born in 1979， Ph. D.， research fellow. Her research interests include industrial artificial intelligence.
LYU Fengfeng， born in 2000， M. S. candidate. Her research interests include bearing life， knowledge graph.
NIU Wei， born in 1979， M. S.， senior engineer. His research interests include enterprise digitalization， intelligent research and development and application.
WU Jian， born in 1984， senior engineer. His research interests include measurement and control， equipment fault diagnosis.
ZHANG Xiao， born in 1990， M. S. His research interests include industrial automation.
Supported by:
Scientific Research Special Project of China National Machinery Industry Corporation(ZDZX2023-11);Youth Science Research Fund of China Machinery Research Industry Corporation Institute(TD2022ZK007);Tianjin Manufacturing Industry High-Quality Development Special Fund(20232181);Hebei Province Higher Education Science and Technology Research Project(CXY2024023)

基于动态知识嵌入的轴承寿命预测方法

刘晶¹^,², 吕凤凤², 牛巍³, 季海鹏¹^,⁴(), 吴健⁵, 张啸⁵

^1.高性能轧辊材料与复合成形全国重点实验室（天津大学），天津 300072
^2.河北工业大学人工智能与数据科学学院，天津 300401
^3.中信戴卡股份有限公司，河北秦皇岛 066011
^4.河北工业大学材料科学与工程学院，天津 300401
^5.天津电气科学研究院有限公司，天津 300180

通讯作者: 季海鹏
作者简介:刘晶（1979—），女，内蒙古包头人，研究员，博士，CCF杰出会员，主要研究方向：工业人工智能
吕凤凤（2000—），女，江苏徐州人，硕士研究生，主要研究方向：轴承寿命、知识图谱
牛巍（1979—），男，黑龙江尚志人，高级工程师，硕士，主要研究方向：企业数字化、智能化研发与应用
吴健（1984—），男，天津人，高级工程师，主要研究方向：测控、设备故障诊断
张啸（1990—），男，河南焦作人，硕士，主要研究方向：工业自动化。
基金资助:
中国机械工业集团有限公司科研专项(ZDZX2023-11);国机研究院青年科研基金资助项目(TD2022ZK007);天津市制造业高质量发展专项资金资助项目(20232181);河北省高等学校科学技术研究项目(CXY2024023)

Abstract

Abstract:

In practical industrial scenarios， the significant divergence in bearing state evolution caused by complex operating conditions imposes dual constraints on bearing life prediction： incomplete mechanism understanding and distribution shift in monitoring data. Although the existing data-driven methods perform well under stable conditions， they suffer from strong reliance on annotated data and limited generalization capability under rappidly changing conditions. To address these issues，a method for Bearing Life Prediction based on Dynamic Knowledge embedding （DK-BLP） was proposed in this paper. The method dynamically integrated data-driven features and domain prior knowledge. Firstly， a dynamic knowledge graph was constructed， the prior knowledge of bearing degradation was encoded into computable triplets， and a sliding window confidence mechanism was introduced， so as to realize adaptive updating of domain knowledge. Second， relational graph convolutional networks were used to extract physically meaningful embedding vectors， which were then fused cross-modally with time-frequency features of vibration signals extracted by hierarchical convolutional networks. Finally， dynamic interactions between features and knowledge were modeled by using a multi-head self-attention Transformer， thereby allowing the model to adaptively balance the contributions of data features and mechanistic knowledge. Experimental results on the PHM2012 and XJTU datasets demonstrate that the proposed method significantly improves the accuracy of cross-condition bearing life prediction.

Key words: knowledge graph, life prediction, domain transfer, knowledge embedding

摘要：

在实际工业场景中，由于复杂工况导致轴承状态演化差异显著，使得轴承寿命预测面临机制认知不完整与监测数据分布漂移的双重约束。现有的基于数据驱动的方法虽在稳定工况下表现良好，但存在标注数据依赖性强和对突变工况泛化能力不足的局限。针对上述问题，本文提出一种基于动态知识嵌入的轴承寿命预测方法（DK-BLP）。该方法动态地融合数据驱动特征与领域先验知识：首先，构建动态知识图谱，将轴承退化的先验知识编码为可计算的三元组知识，并引入滑动窗口置信度机制，实现领域知识的自适应更新；其次，利用关系图卷积网络提取具有物理意义的嵌入向量，并把它们与层级卷积网络提取的振动信号时频特征进行跨模态融合；最后，通过多头自注意力Transformer实现特征与知识的动态关联建模，使模型能够自适应地平衡数据特征与机制知识的贡献。在PHM2012和XJTU数据集上的实验结果表明，本文方法显著提升了跨工况的轴承寿命预测精度。

关键词: 知识图谱, 寿命预测, 领域迁移, 知识嵌入

CLC Number:

TP391.4TP301.6

Jing LIU, Fengfeng LYU, Wei NIU, Haipeng JI, Jian WU, Xiao ZHANG. Bearing life prediction method based on dynamic knowledge embedding[J]. Journal of Computer Applications, 2026, 46(7): 2347-2354.

刘晶, 吕凤凤, 牛巍, 季海鹏, 吴健, 张啸. 基于动态知识嵌入的轴承寿命预测方法[J]. 《计算机应用》唯一官方网站, 2026, 46(7): 2347-2354.

Figures/Tables 13

References 21

[1]	Sun B， Hu W， Wang H， et al. Remaining useful life prediction of rolling bearings based on CBAM-CNN-LSTM ［J］. Sensors， 2025， 25（2）： No.554.
[2]	Yang L， Jiang Y， Zeng K， et al. Rolling bearing remaining useful life prediction based on CNN-VAE-MBiLSTM ［J］. Sensors， 2024， 24（10）： No.2992.
[3]	刘晶，董志红，张喆语，等.基于联邦增量学习的工业物联网数据共享方法［J］.计算机应用， 2022， 42（4）： 1235-1243.
	Liu Jing， Dong Zhihong， Zhang Zheyu， et al. Data sharing method of industrial internet of things based on federal incremental learning ［J］. Journal of Computer Applications， 2022， 42（4）： 1235-1243.
[4]	Zhu W， Ni G， Cao Y， et al. Research on a rolling bearing health monitoring algorithm oriented to industrial big data ［J］. Measurement， 2021， 185： No.110044.
[5]	Hou W， Peng Y. Enhancing bearing life prediction： sparse Gaussian process regression approach based on sequential ensemble and residual reduction for degradation prediction ［J］. Reliability Engineering and System Safety， 2025， 256： No.110788.
[6]	Wei Y， Wu D. Conditional variational Transformer for bearing remaining useful life prediction ［J］. Advanced Engineering Informatics， 2024， 59： No.102247.
[7]	文井辉，伍荣森，李帅永，等.基于DRSN和优化BiLSTM的轴承剩余寿命预测方法［J］.计算机集成制造系统， 2024， 30（5）： 1877-1888.
	Wen Jinghui， Wu Rongsen， Li Shuaiyong， et al. Bearing residual life prediction method based on DRSN and optimized BiLSTM ［J］. Computer Integrated Manufacturing Systems， 2024， 30（5）： 1877-1888.
[8]	Yang X， Zheng Y， Zhang Y， et al. Bearing remaining useful life prediction based on regression Shapalet and graph neural network ［J］. IEEE Transactions on Instrumentation and Measurement， 2022， 71： No.3505712.
[9]	Wang H， An J， Yang J， et al. Remaining useful life prediction method of bearings based on the interactive learning strategy ［J］. Computers and Electrical Engineering， 2025， 121： No.109853.
[10]	He D， Zhao J， Jin Z， et al. DCAGGCN： a novel method for remaining useful life prediction of bearings ［J］. Reliability Engineering and System Safety， 2025， 260： No.110978.
[11]	Yang L， Li T， Dong Y， et al. A knowledge-data integration framework for rolling element bearing RUL prediction across its life cycle ［J］. ISA Transactions， 2024， 152： 331-357.
[12]	Li C， Zhai W， Fu W， et al. Remaining useful life prediction of rolling bearings based on parallel feature extraction ［J］. Robotic Intelligence and Automation， 2024， 45（1）： 90-105.
[13]	Zhan F， Hu L K， Huang W K， et al. Category knowledge-guided few-shot bearing fault diagnosis ［J］. Engineering Applications of Artificial Intelligence， 2025， 139： No.109489.
[14]	Peng C， Sheng Y， Gui W， et al. A rolling bearing fault diagnosis method based on multimodal knowledge graph ［J］. IEEE Transactions on Industrial Informatics， 2024， 20（11）： 13047-13057.
[15]	Ma L， Jiang B， Lu N， et al. Aeroengine bearing time-varying skidding assessment with prior knowledge-embedded dual feedback spatial-temporal GCN ［J］. IEEE Transactions on Cybernetics， 2025， 55（2）： 826-839.
[16]	Nectoux P， Gouriveau R， Medjaher K， et al. PRONOSTIA： an experimental platform for bearings accelerated life test ［C］// PHM 2012. Piscataway： IEEE， 2012：1-8.
[17]	Wang B， Lei Y， Li N， et al. A hybrid prognostics approach for estimating remaining useful life of rolling element bearings ［J］. IEEE Transactions on Reliability， 2020， 69（1）： 401-412.
[18]	高萌，鲁玉军. 基于Bi-TCN-LSTM的滚动轴承剩余使用寿命预测方法［J］. 轻工机械， 2024， 42（3）： 66-73， 79.
	Gao Meng， Lu Yujun. Prediction method for remaining useful life of rolling bearings based on bidirectional temporal convolutional network and long short-term memory network［J］. Light Industry Machinery， 2024， 42（3）： 66-73， 79.
[19]	史竞成，吴占涛，程军圣，等. 改进自注意力机制的滚动轴承寿命预测方法［J］. 噪声与振动控制， 2025， 45（2）： 90-96， 104.
	Shi Jingcheng， Wu Zhantao， Cheng Junsheng， et al. Residual life prediction method of rolling bearings based on improved self-attention mechanism［J］. Noise and Vibration Control， 2025， 45（2）： 90-96， 104.
[20]	李广福，马萍，张宏立，等. 一种Informer模型的滚动轴承剩余寿命预测方法［J］. 机械科学与技术， 2024， 43（12）： 2016-2023.
	Li Guangfu， Ma Ping， Zhang Hongli， et al. A method for predicting remaining life of rolling bearings using Informer model［J］. Mechanical Science and Technology for Aerospace Engineering， 2024， 43（12）： 2016-2023.
[21]	韩延，林志超，黄庆卿，等. 正交约束域适应的跨工况滚动轴承剩余使用寿命预测方法［J］. 电子与信息学报， 2024， 46（3）： 1043-1050.
	Han Yan， Lin Zhichao， Huang Qingqing， et al. A domain adaptive method with orthogonal constraint for predicting the remaining useful life of rolling bearings under cross working conditions［J］. Journal of Electronics and Information Technology， 2024， 46（3）： 1043-1050.

工况	转速/（r·min^-1）	负载/N	训练集
1	1 800	4 000	1_1，1_2
2	1 650	4 200	2_1，2_2
3	1 500	5 000	3_1，3_2

工况	转速/（r·min^-1）	负载/N	训练集
1	1 800	4 000	1_1，1_2
2	1 650	4 200	2_1，2_2
3	1 500	5 000	3_1，3_2

工况	转速/（r·min^-1）	负载/N	验证集
1	2 100	12 000	1_1，1_2，1_3，1_4，1_5
2	2 250	11 000	2_1，2_2，2_3，2_4，2_5
3	2 400	10 000	3_1，3_2，3_3，3_4，3_5

工况	转速/（r·min^-1）	负载/N	验证集
1	2 100	12 000	1_1，1_2，1_3，1_4，1_5
2	2 250	11 000	2_1，2_2，2_3，2_4，2_5
3	2 400	10 000	3_1，3_2，3_3，3_4，3_5

样本	指标	Informer	Transformer	Bi-TCN-LSTM	MCD-OC	DK-BLP
1	MAE	0.203 5	0.262 9	0.475 0	0.229 8	0.079 5
	RMSE	0.232 3	0.304 1	0.547 3	0.263 6	0.110 7
	MSE	0.053 9	0.092 5	0.299 5	0.069 5	0.012 3
2	MAE	0.339 9	0.367 7	0.447 8	0.258 0	0.193 3
	RMSE	0.400 9	0.448 9	0.516 7	0.311 5	0.237 3
	MSE	0.160 7	0.201 5	0.266 9	0.097 0	0.056 3
3	MAE	0.752 0	0.306 9	0.401 4	0.206 7	0.134 7
	RMSE	0.822 8	0.397 9	0.481 1	0.250 5	0.212 2
	MSE	0.677 0	0.158 3	0.231 5	0.062 8	0.045 0
4	MAE	0.319 6	0.336 3	0.472 2	0.160 1	0.129 1
	RMSE	0.363 3	0.400 7	0.548 8	0.188 8	0.157 4
	MSE	0.132 0	0.160 6	0.301 2	0.035 6	0.024 8
5	MAE	0.195 4	0.349 9	0.382 6	0.250 1	0.105 3
	RMSE	0.240 8	0.434 7	0.453 0	0.301 6	0.127 2
	MSE	0.058 0	0.189 0	0.205 2	0.090 9	0.016 2
6	MAE	0.214 1	0.275 6	0.448 1	0.232 3	0.129 9
	RMSE	0.249 0	0.343 7	0.528 8	0.281 7	0.168 9
	MSE	0.062 0	0.118 2	0.279 6	0.079 3	0.028 5

Bearing life prediction method based on dynamic knowledge embedding

基于动态知识嵌入的轴承寿命预测方法

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 13

References 21

Related Articles 15

Recommended Articles

Metrics

[1]	Jiahang LI, Qilong HAN, Lijie LI, Hui ZHANG. Large language model-driven method for qualifier enhancement in hyper-relational knowledge graphs [J]. Journal of Computer Applications, 2026, 46(7): 2043-2053.
[2]	Meihua WANG, Jie HUANG, Wen WEN, Ruichu CAI, Peijie HUANG, Yuhong XU, Xinlong LIN. Gradient orthogonal projection based continual embedding method for dynamic knowledge graph [J]. Journal of Computer Applications, 2026, 46(6): 1776-1784.
[3]	Ronghui ZHAO, Chao DENG, Zidong YU. Construction and application of knowledge graph for fault diagnosis of key components of aviation equipment [J]. Journal of Computer Applications, 2026, 46(5): 1604-1613.
[4]	Hao LIANG, Shaojie QIAO. Complex query-based question-answering model integrating bidirectional sequence embeddings [J]. Journal of Computer Applications, 2026, 46(4): 1096-1103.
[5]	Kaizhou SHI, Xuan HE, Guoyi HOU, Gen LI, Shuanggao LI, Xiang HUANG. Airborne product metrological traceability knowledge graph construction method based on large language models [J]. Journal of Computer Applications, 2026, 46(4): 1086-1095.
[6]	Haoyang ZHANG, Liping ZHANG, Sheng YAN, Na LI, Xuefei ZHANG. Review of large language model methods for knowledge graph completion [J]. Journal of Computer Applications, 2026, 46(3): 683-695.
[7]	Yiming HUANG, Xihua ZOU, Guo DENG, Di ZHENG. Pre-answering and retrieval filtering： dual-stage optimization method for RAG-based question-answering systems [J]. Journal of Computer Applications, 2026, 46(3): 696-707.
[8]	Xue WANG, Liping ZHANG, Sheng YAN, Na LI, Xuefei ZHANG. Review of multi-modal knowledge graph completion methods [J]. Journal of Computer Applications, 2026, 46(2): 341-353.
[9]	Jindong HE, Yuxuan JI, Tianci CHEN, Hengming XU, Ji GENG, Mingsheng CAO, Yuanning LIANG. Entity discovery method for non-intelligent sensors by integrating knowledge graph and large models [J]. Journal of Computer Applications, 2026, 46(2): 354-360.
[10]	Fei WANG, Ye TAO, Jiawang LIU, Wei LI, Xiugong QIN, Ning ZHANG. Bimodal fusion method for constructing spatio-temporal knowledge graph in smart home space [J]. Journal of Computer Applications, 2026, 46(1): 52-59.
[11]	Yanan LI, Mengyang GUO, Guojun DENG, Yunfeng CHEN, Jianji REN, Yongliang YUAN. Method for life prediction of parallel branching engine based on multi-modal fusion features [J]. Journal of Computer Applications, 2026, 46(1): 305-313.
[12]	Chao LIU, Yanhua YU. Knowledge-aware recommendation model combining denoising strategy and multi-view contrastive learning [J]. Journal of Computer Applications, 2025, 45(9): 2827-2837.
[13]	Yuyang SUN, Minjie ZHANG, Jie HU. Zero-shot dialogue state tracking domain transfer model based on semantic prefix-tuning [J]. Journal of Computer Applications, 2025, 45(7): 2221-2228.
[14]	Shuang LIU, Daqing LIU, Jiana MENG, Di ZHAO. Hyper-relational knowledge graph completion method fusing noise filtering [J]. Journal of Computer Applications, 2025, 45(6): 1817-1826.
[15]	Sheping ZHAI, Yan HUANG, Qing YANG, Rui YANG. Multi-view entity alignment combining triples and text attributes [J]. Journal of Computer Applications, 2025, 45(6): 1793-1800.