Aiming at the limitations of traditional graph convolutional networks (GCNs) in dynamic feature capture and multi-dimensional feature fusion, this paper proposes a novel method called multi-level adaptive fusion and edge-enhanced attention GCN (ME-GCN). The ME-GCN integrates the multi-level dynamic collaborative fusion mechanism, which combines feature-space and topology-space graph convolution modules, enabling comprehensive extraction of local features and global topological relations. By introducing the edge-enhanced graph attention mechanism, the association weights between nodes are dynamically adjusted, thereby enhancing the modeling capability of complex fault patterns. Moreover, the adaptive fusion strategy dynamically balances and optimizes multi-module features, achieving efficient and robust fault representation. Experimental validation on both public datasets and real industrial datasets demonstrates that ME-GCN achieves superior diagnostic performance with strong robustness and generalization ability under different operating conditions.
LiXWuXYuJ, et al. A fault diagnosis data augmentation method integrating multimodal non-Gaussian denoising diffusion generative adversarial network. Adva Eng Inform2025; 68: 103776. https://doi.org/10.1016/j.aei.2025.103776
5.
YanSShaoHLongZ, et al. Incremental few-shot fault diagnosis with cosine-represent real-time evolved network of transmission systems. IEEE Transactions on Transportation Electrification, 2025; 11(4): 9544–9554. https://doi.org/10.1109/TTE.2025.3563630
6.
ZhiSNiuYMaL, et al. Local entropy selection scaling-extracting chirplet transform for enhanced timefrequency analysis and precise state estimation in reliability-focused fault diagnosis of non-stationary signals. Maint Reliab2026; 28(1): 205977. https://doi.org/10.17531/ein/205977
7.
ZhuangJWuJPeiG, et al. A neuro-fuzzy approach with hypergraph convolution for fault diagnosis in industrial devices. J Reliab Sci Eng2025; 1(3): 035301. https://doi.org/10.1088/3050-2454/ae064d
8.
ShaoHMingYLiuY, et al. Small sample gearbox fault diagnosis based on improved deep forest in noisy environments. Nondestr Test Eval2025; 40(8): 3935–3956. https://doi.org/10.1080/10589759.2024.2404489
9.
YaoZCuiXGebraeelN, et al. Metric learning-based two-stage imbalanced fault diagnosis model for water injection pump. J Reliab Sci Eng2025; 1(4): 045402. https://doi.org/10.1088/3050-2454/ae227a
10.
GuoYLiSXiaT, et al. A novel cross-domain open-set fault diagnosis model for distribution boxes in finishing rolling group. J Reliab Sci Eng2026; 2(1): 015302. https://doi.org/10.1088/3050-2454/ae4e31
11.
ZhiSSuKYuJ, et al. An unsupervised transfer learning bearing fault diagnosis method based on multi-channel calibrated Transformer with shiftable window. Struct Health Monit2025; 0(0). https://doi.org/10.1177/14759217251324671
12.
ZhiSWuHShenH, et al. Transparent operator network: a fully interpretable network incorporating learnable wavelet operator for intelligent fault diagnosis. IEEE Trans Ind Inform2024; 20(6): 8628–8638. https://doi.org/10.1109/TII.2024.3366993
13.
LinHHuangXChenZ, et al. Matching pursuit network: an interpretable sparse time–frequency representation method toward mechanical fault diagnosis. IEEE Trans Neural Netw Learn Syst2024; 36(7): 12377–12388. https://doi.org/10.1109/TNNLS.2024.3483954
14.
LiXYuanPWangX, et al. An unsupervised transfer learning bearing fault diagnosis method based on depthwise separable convolution. Meas Sci Technol2023; 34(9): 095401. https://doi.org/10.1088/1361-6501/acda55
15.
BaXZhangXLiS, et al. Multimodal semantic communication system based on graph neural networks. Intell Robot2025; 5(3): 805–826. https://doi.org/10.20517/ir.2025.41
16.
WuZPanSChenF, et al. A comprehensive survey on graph neural networks. IEEE Trans Neural Netw Learn Syst2020; 32(1): 4–24.
17.
MaZGongH. Heterogeneous multi-agent task allocation based on graph neural network ant colony optimization algorithms. Intell Robot2023; 3(4): 581–595. https://doi.org/10.20517/ir.2023.33
18.
ZhangSTongHXuJ, et al. Graph convolutional networks: a comprehensive review. Computat Soc Netw2019; 6(1): 1–23.
19.
WangXZhouWLiX. Bearing fault diagnosis based on graph formulation and graph convolutional network. J Dyn Monit Diagn2023; 2(4): 252–261. https://doi.org/10.37965/jdmd.2023.468
20.
WangXZhouDZhuangX, et al. SGG-DGCN: wind turbine anomaly identification by using deep graph convolutional networks with similarity graph generation strategy. J Dyn Monit Diagn2024; 3(4): 258–267. https://doi.org/10.37965/jdmd.2024.558
21.
VeličkovićPCucurullGCasanovaA, et al. Graph attention networks. In: International Conference on Learning Representations, 2018: Ithaca.
22.
LiXLiGWangT, et al. Knowledge extraction and retrieval-augmented generation for intelligent maintenance of wind power equipment based on graph attention networks. Chin J Mech Eng2025; 100141. https://doi.org/10.1016/j.cjme.2025.100141
23.
QiJChenZKongY, et al. Attention-guided graph isomorphism learning: a multi-task framework for fault diagnosis and remaining useful life prediction. Reliab Eng Syst Saf2025; 263: 111209. https://doi.org/10.1016/j.ress.2025.111209
24.
LiXRanGXiaoS, et al. A fault diagnosis method for bearings based on multiscale graph convolutional network under non-stationary speed conditions. Mech Mach Theory2025; 217: 106267. https://doi.org/10.1016/j.mechmachtheory.2025.106267
HuoGZhangYGaoJ, et al. CaEGCN: cross-attention fusion based enhanced graph convolutional network for clustering. IEEE Trans Knowl Data Eng2021; 35(4): 3471–3483. https://doi.org/10.1109/TKDE.2021.3125020
27.
LiXXiaoSLiQ, et al. The bearing multi-sensor fault diagnosis method based on a multi-branch parallel perception network and feature fusion strategy. Reliab Eng Syst Saf2025; 261: 111122. https://doi.org/10.1016/j.ress.2025.111122
KhemaniBPatilSKotechaK, et al. A review of graph neural networks: concepts, architectures, techniques, challenges, datasets, applications, and future directions. J Big Data2024; 11(1): 18. https://doi.org/10.1186/s40537-023-00876-4
30.
SteinbachMTanP-N. kNN: k-nearest neighbors. In: WuXKumarV(eds) The top ten algorithms in data mining. Boca Raton, FL: Chapman and Hall/CRC, 2009, pp. 165–176.
31.
ChenZXuJPengT, et al. Graph convolutional network-based method for fault diagnosis using a hybrid of measurement and prior knowledge. IEEE Trans Cyber2021; 52(9): 9157–9169. https://doi.org/10.1109/TCYB.2021.3059002
32.
KolouriSNaderializadehNRohdeGK, et al. Wasserstein embedding for graph learning. In: International Conference on Learning Representations, Vienna, Austria, 2021.
33.
YuYWangZLiaoY, et al. Wasserstein distance-based graph contrastive learning for recommendation. Expert Syst Appl2025; 129427. https://doi.org/10.1016/j.eswa.2025.129427
34.
LuZ-YZhuQ-XHeY-L. Wasserstein distance-enhanced variable edge-weight graph convolutional networks for industrial fault diagnosis. ISA Trans2025; 167: 78–89.
35.
McInnesLHealyJSaulN, et al. Umap: uniform manifold approximation and projection for dimension reduction. J Open Source Softw2018; 3(29): 861.
36.
BrodySAlonUYahavE. How attentive are graph attention networks? In: International Conference on Learning Representations, Virtual Event, 2022.
37.
HamiltonWYingZLeskovecJ. Inductive representation learning on large graphs. Adv Neural Inform Process Syst2017; 30: 1024–1034.
38.
WangXZhuMBoD, et al. Am-gcn: adaptive multi-channel graph convolutional networks. In: Proceedings of the 26th ACM SIGKDD International conference on knowledge discovery & data mining. Virtual Event, CA, USA. New York, NY, USA: Association for Computing Machinery, 23–27 August 2020, pp. 1243–1253.
39.
Van der MaatenLHinton. Visualizing data using t-SNE. J Mach Learn Res2008; 9(Nov): 2579–2605.
40.
HuangRXiaJZhangB, et al. Compound fault diagnosis for rotating machinery: state-of-the-art, challenges, and opportunities. J Dyn Monit Diagn2023; 2(1): 13–29. https://doi.org/10.37965/jdmd.2023.152
41.
SmithWARandallRB. Rolling element bearing diagnostics using the Case Western Reserve University data: a benchmark study. Mech Syst Signal Process2015; 64: 100–131. https://doi.org/10.1016/j.ymssp.2015.04.021
42.
RanGLiXLiM, et al. FD-Zero framework: LLM-based diagnosis framework for zero-shot mechanical time-domain signals. Struct Health Monit2025; 0(0). https://doi.org/10.1177/14759217251366940
43.
ChenJBenestyJHuangY, et al. New insights into the noise reduction Wiener filter. IEEE Trans Audio Speech Lang Process2006; 14(4): 1218–1234. https://doi.org/10.1109/TSA.2005.860851
44.
ShaoHLaiYLiuH, et al. LSFConvformer: a lightweight method for mechanical fault diagnosis under small samples and variable speeds with time-frequency fusion. Mech Syst Signal Process2025; 236: 113016. https://doi.org/10.1016/j.ymssp.2025.113016
45.
LiXLiSZhangG, et al. A novel interpretable dynamic weighted domain adaptation network for cross-domain fault diagnosis of bearings under time-varying speeds. Eng Appl Artif Intell2026; 170: 114173. https://doi.org/10.1016/j.engappai.2026.114173
