Sample-optimized adaptive perceptual enhanced graph neural network based fault identification in industrial processes

Yao Keyu; Shi Hongbo; Song Bing; Tao Yang

doi:null

Your Location：

Home >

Browse articles >

Sample-optimized adaptive perceptual enhanced graph neural network based fault identification in industrial processes

Updated：2026-05-14

- Sample-optimized adaptive perceptual enhanced graph neural network based fault identification in industrial processes
- Chinese Journal of Chemical Engineering Vol. 90, Issue 2, Pages: 181-195(2026)
- Affiliations：
  
  Key Laboratory of Smart Manufacturing in Energy Chemical Process, Ministry of Education, East China University of Science and Technology, Shanghai 200237, China
- Author bio：
  
  Corresponding author. East China University of Science and Technology, Shanghai 200237, China. E-mail address: hbshi@ecust.edu.cn (H. Shi).
- Funds：
- DOI：
  CLC：
- Received：30 August 2024，
  
  Revised：2025-09-08，
  
  Accepted：20 October 2025，
  
  Online First：31 October 2025，
  
  Published：2026-02
- Accepted：
Scan QR Code
Yao Keyu, Shi Hongbo, Song Bing, et al. Sample-optimized adaptive perceptual enhanced graph neural network based fault identification in industrial processes[J]. Chinese Journal of Chemical Engineering, 2026, 90(2): 181-195.
DOI：

Yao Keyu, Shi Hongbo, Song Bing, et al. Sample-optimized adaptive perceptual enhanced graph neural network based fault identification in industrial processes[J]. Chinese Journal of Chemical Engineering, 2026, 90(2): 181-195. DOI：

摘要

Abstract

With the complexity and intelligence of the industrial process

the identification of faults in the actual process plays a crucial role in ensuring production safety. The traditional fault identification strategies have the problem that similar characterized faults are unable to be accurately identified. Motivated by the limitations

a novel sample-optimized adaptive perceptual enhanced graph neural network (SOAP-EGNN) for large-scale process fault identification is proposed. Initially

process mechanism knowledge and process data correlation are injected into the modeling approach through graph neural networks

and the transmission of information based on the enhanced attention mechanism is introduced to describe the quantitative relationships between process variables at a fine-grained level based on the adaptive perception strategy. Subsequently

to achieve better intra-class compactness and inter-class separability in feature representation

our designed sample-optimized feature processing strategy (SOFPS) is applied. Furthermore

to enhance the robustness and generalization capability of the model during training

a label smoothing regularization (LSR) strategy is incorporated. This approach effectively mitigates the risk of overfitting by introducing a degree of uncertainty into the label space

thereby encouraging the model to learn more discriminative and stable features. Ultimately

the efficacy and superiority of theSOAP-EGNNalgorithm are thoroughly validated through comprehensive simulation experiments conducted on the Tennessee Eastman process (TEP).

关键词

Keywords

references

W.D. Tian, J.W. Zhang, Z. Cui, H.R. Zhang, B. Liu, Microscopic mechanism study and process optimization of dimethyl carbonate production coupled biomass chemical looping gasification system, Chin. J. Chem. Eng. 58 (2023) 291—305.

F.N. Irani, M. Yadegar, N. Meskin, Koopman-based deep iISS bilinear parity approach for data-driven fault diagnosis: experimental demonstration using three-tank system, Control Eng. Pract. 142(2024)105744.

S.W. Xiong, L. Zhou, Y.Y. Dai, X. Ji, Attention-based long short-term memory fully convolutional network for chemical process fault diagnosis, Chin. J. Chem. Eng. 56 (2023)1—14.

C.H. Zhao, Perspectives on nonstationary process monitoring in the era of industrial artificial intelligence, J. Process Control 116 (2022)255—272.

W.D. Tian, G.X. Zhang, X. Zhang, Y.X. Dong, PCA weight and Johnson transformation based alarm threshold optimization in chemical processes, Chin. J. Chem. Eng. 26 (2018) 1653—1661.

A.N. M, S. R, Combined classification models for bear ing fault diagnosis with improved ICA and MFCC feature set, Adv. Eng. Software 173(2022)103249.

K.X. Bi, T. Qiu, An intelligent SVM modeling process for crude oil properties prediction based on a hybrid GA-PSO method, Chin. J. Chem. Eng. 27 (2019) 1888—1894.

N.X. Li, H.B. Shi, B. Song, Y. Tao, Temporal-spatial neighborhood enhanced sparse autoencoder for nonlinear dynamic process monitoring, Processes 8 (2020)1079.

C. Yang, Z. Lai, Y.C. Wang, S.L. Lan, L.H. Wang, L.H. Zhu, A novel bearing fault diagnosis method based on stacked autoencoder and end-edge collaboration, in:2023 26th Int. Conf. Comput. Support. Coop. Work Des. CSCWD, IEEE, Rio de Janeiro, Brazil,2023.

S.M. Chen, J.B. Yu, S.J. Wang, One-dimensional convolutional auto-encoderbased feature learning for fault diagnosis of multivariate processes, J. Process Control 87 (2020) 54—67.

J.X. Zhang, M. Zhang, Z.M. Feng, R.F. Lv, C.Y. Lu, Y.Y. Dai, L.C. Dong, Gated recurrent unit-enhanced deep convolutional neural network for real-time industrial process fault diagnosis, Process Saf. Environ. Prot. 175 (2023) 129—149.

Z.H. Wu, S.R. Pan, F.W. Chen, G.D. Long, C.Q. Zhang, P.S. Yu, A comprehensive survey on graph neural networks, IEEE Transact. Neural Networks Learn. Syst. 32(2021)4—24.

F. Xia, K. Sun, S. Yu, A. Aziz, L. Wan, S. Pan, H. Liu, Graph learning: a survey, IEEE Trans. Artif. Intell. 2(2021)109—127.

C.Y. Yang, J. Liu, K.B. Zhou, X.H. Yuan, M.-F. Ge, Transfer graph-driven rotating machinery diagnosis considering cross-domain relationship construction, IEEE ASME Trans. Mechatron. 27(2022)5351—5360.

T.F. Li, Z. Zhou, S.N. Li, C. Sun, R.Q. Yan, X.F. Chen, The emerging graph neural networks for intelligent fault diagnostics and prognostics: a guideline and a benchmark study, Mech. Syst. Signal Process. 168 (2022) 108653.

D. Wu, J.S. Zhao, Process topology convolutional network model for chemical process fault diagnosis, process, Saf. Environ. Prot. 150 (2021) 93—109.

W.Q. Wu, C.Y. Song, J. Liu, J. Zhao, Data-knowledge-driven distributed monitoring for large-scale processes based on digraph, J. Process Control 109 (2022)60—73.

Z.C. Chen, Z.Q. Ge, Knowledge automation through graph mining, convolution, and explanation framework: a soft sensor practice, IEEE Trans. Ind. Inf. 18 (9) (2021) 6068—6078.

Q. Deng, Z.W. Chen, P. Tang, X.H. Li, C.H. Yang, X. Yang, A novel semisupervised fault diagnosis method for chillers based on neighboroptimized graph convolutional network, Energy Build. 301 (2023) 113703.

A. Bathelt, N.L. Ricker, M. Jelali, Rev ision of the Tennessee eastman process model, IFAC-PapersOnLine 48(2015)309—314.

S. Yin, S.X. Ding, A. Haghani, H.Y. Hao, P. Zhang, A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process, J. Process Control 22 (9) (2012) 1567—1581.

Q.C. Jiang, X.F. Yan, B. Huang, Performance-driven distributed PCA process monitoring based on fault-relevant variable selection and bayesian inference, IEEE Trans. Ind. Electron. 63 (2016).

Y.M. He, H.B. Shi, S. Tan, J.Z. Zhu, Multiblock temporal convolution networkbased temporal-correlated feature learning for fault diagnosis of multivariate processes, J. Taiwan Inst. Chem. Eng. 122 (2021) 78—84.

S.Y. Liu, J. He, Z.W. Chen, D.F. Chen, Y. Chen, Discriminative stacked autoencoder: feature-integration boosting for bearing fault diagnosis, IEEE Sens. J. 23(2023)27549—27558.

S.P. Gao, Y.Y. Li, D. Zhao, Fault diagnosis for rolling bearings based on novel visibility graph and GCN scheme, in: 2023 IEEE 12th Data Driven Control Learn. Syst. Conf. DDCLS, IEEE, Xiangtan, China,2023.

Y.D. Xu, J.C. Ji, Q. Ni, K. Feng, M. Beer, H.T. Chen, A graph-guided collaborative convolutional neural network for fault diagnosis of electromechanical systems, Mech. Syst. Signal Process. 200 (2023) 110609.

D.Y. Chen, R.N. Liu, Q.H. Hu, S.X. Ding, Interaction-aware graph neural networks for fault diagnosis of complex industrial processes, IEEE Transact. Neural Networks Learn. Syst. 34 (9) (2021) 6015—6028.

H. Wang, R.N. Liu, Z.X. Li, H.K. Zhou, Causal-trivial attention graph neural network for fault diagnosis of complex industrial processes, IEEE Trans. Ind. Inf. 20 (2)(2023)1987—1996.

B. Wang, Y.D. Xu, M.Y. Wang, Y.Z. Li, Gear fault diagnosis method based on the optimized graph neural networks, IEEE Trans. Instrum. Meas. 73 (2023) 1—11.

T.-N. Hoang, D. Kim, Supervised contrastive ResNet and transfer learning for the in-vehicle intrusion detection system, Expert Syst. Appl. 238 (2024) 122181.

J. Wan, D.X. Zhong, H.K. Shao, Palmprint recognition system for mobile device based on circle loss, Displays 73 (2022) 102214.

C.C. Fan, Y.X. Wang, Y.H. Zhang, W.L. Ouyang, Interpretable multi-scale neural network for granger causality discovery, in: ICASSP 2023-2023 IEEE Int. Conf. Acoust. Speech Signal Process. ICASSP, IEEE, Rhodes Island, Greece, 2023.

R.C. Zhi, Y.C. Meng, J.Y. Hou, J. Wan, Dual balanced class-incremental learning with im-Softmax and angular rectification, IEEE Transact. Neural Networks Learn. Syst. 36(3)(2025) 4437—4447.

X.C. Zhan, A.M. Li, J. He, L. Xiang, Enhancing low-dose pet imaging: a novel contrastive learning method for perceptual loss and an organ-aware loss, in:2024 IEEE International Symposium on Biomedical Imaging, IEEE, 2024.

K.A. Cahyanto, M. Yani, A. Kustiyo, The effects of L2 regularization in EfficientNet for human skin disease multi-class classification, in: 2023 6th International Conference of Computer and Informatics Engineering, IEEE, 2023.

Z. Liu, H.J. Gu, Z. Chen, B. Pierre, Y.J. Yang, Dual-modal image reconstruction for electrical impedance tomography with overlapping group lasso and Laplacian regularization, IEEE Trans. Biomed. Eng. 70(8)(2023) 2362—2373.

X.X. Liu, C.C. Zou, C. Peng, C.W. Wu, Uncertainty quantification in intelligentbased electrical resistivity tomography image reconstruction with Monte Carlo dropout strategy, IEEE Trans. Geosci. Rem. Sens. 61 (2023) 1—16.

Views

Downloads

CSCD

Alert me when the article has been cited

Submit

Tools

Publicity Resources

Hierarchical multihead self-attention for time-series-based fault diagnosis

Hybrid modelling incorporating reaction and mechanistic data for accelerating the development of isooctanol oxidation

Knowledge graphs in heterogeneous catalysis: Recent advances and future opportunities

SmdaNet: A hierarchical hard sample mining and domain adaptation neural network for fault diagnosis in industrial process

Consequence prediction using variable-length concentration time series for gas turbine enclosure

Related Author

Chengtian Wang

Hongbo Shi

Bing Song

Yang Tao

Xin Zhou

Ce Liu

Zhibo Zhang

Xinrui Song

Related Institution

Key Laboratory of Smart Manufacturing in Energy Chemical Process of the Ministry of Education, East China University of Science and Technology

State Key Laboratory of Heavy Oil Processing, China University of Petroleum

Department of Chemistry and Chemical Engineering, Ocean University of China

Virginia Polytechnic Institute and State University, Blacksburg

SINOPEC Research Institute of Safety Engineering Co., Ltd

⁰