Spatiotemporal Graph Neural Networks for Fault Detection and Structural Learning in Chemical Processes: Use Case on the Tennessee Eastman Process
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
https://orcid.org/0000-0002-4637-6826
Victor COSTA Fanny CASTERAN Benoit CELSE Benoit IUNG
Abstract
Classical Fault Detection and Diagnosis (FDD) methods, including many data-driven approaches, assume a static normal operating space and interpret deviations from a fixed reference as fault indicators. When Operating Conditions (OCs) vary over time, this assumption breaks down: legitimate transitions trigger false alarms while moderate faults go undetected. We propose a spatiotemporal Graph Neural Network (GNN) framework that decomposes the normal operating space into OC-specific subspaces linked by transition functions, with a dual learning objective combining reconstruction loss and a Deep Support Vector Data Description (DeepSVDD) one-class term. The framework learns adjacency matrices through Graph Attention Networks (GATs) and integrates spatial modelling with temporal encoding to represent process dynamics under evolving OCs.
This paper evaluates the foundational components of the framework — fault detection, spatial graph learning via GATv2, and temporal encoding — on the Tennessee Eastman Process (TEP) benchmark, with training performed exclusively on fault-free data. The spatiotemporal architecture achieves competitive detection performance from reconstruction error alone, with similarity-based feature selection improving both accuracy and graph structure diversity. We then evaluate the physical interpretability of the learned attention matrices against 22 ground-truth sensor pairs derived from the TEP control structure and process topology. The GATv2 attention does not recover all the necessary known physical pairs across multiple hyperparameter configurations, suggesting a structural limitation of reconstruction-driven attention rather than a tuning issue. This result challenges a common assumption in GNN-based FDD: that learned attention weights provide a basis for fault diagnosis and root-cause analysis. The architecture detects faults effectively, but the learned graph does not encode the physical topology needed for interpretable diagnosis, motivating physics-informed graph construction.
How to Cite
##plugins.themes.bootstrap3.article.details##
Fault Detection and Diagnosis, Graph Neural Networks, Graph Attention Networks, Spatiotemporal Modelling, Process Monitoring, Tennessee Eastman Process
Brody, S., Alon, U., & Yahav, E. (2022). How attentive are graph attention networks? In Proceedings of the International Conference on Learning Representations (ICLR).
Chen, D., Liu, R., Hu, Q., & Ding, S. X. (2023). Interaction-aware graph neural networks for fault diagnosis of complex industrial processes. IEEE Transactions on Neural Networks and Learning Systems, 34(9), 6015–6028. doi: 10.1109/TNNLS.2021.3132376
Chiang, L. H., Russell, E. L., & Braatz, R. D. (2001). Fault detection and diagnosis in industrial systems. London: Springer.
Deng, A., & Hooi, B. (2021). Graph neural network-based anomaly detection in multivariate time series. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 35, pp. 4027–4035). AAAI Press.
Downs, J. J., & Vogel, E. F. (1993). A plant-wide industrial process control problem. Computers & Chemical Engineering, 17(3), 245–255.
Filonov, P., Lavrentyev, A., & Vorontsov, A. (2016). Multivariate industrial time series with cyber-attack simulation: Fault detection using an LSTM-based predictive data model. In NIPS 2016 Time Series Workshop. arXiv:1612.06676.
Jia, M., Yang, T., Wang, Y., Xu, H., & Liu, B. (2023). Topology-guided graph learning for process fault diagnosis. Industrial & Engineering Chemistry Research, 62(7), 3238–3251. doi: 10.1021/acs.iecr.2c03628
Kipf, T. N., & Welling, M. (2017). Semi-supervised classification with graph convolutional networks. In Proceedings of the International Conference on Learning Representations (ICLR). Toulon, France.
Kovalenko, I., Pozdnyakov, V., & Makarov, I. (2024). GNN with trainable adjacency matrix for fault diagnosis. IEEE Access, 12, 152860–152874. doi: 10.1109/ACCESS.2024.3481331
Liu, Y., & Jafarpour, B. (2024). CNN-GAT with Granger causality for process monitoring. Computers & Chemical Engineering, 180, 108453. doi: 10.1016/j.compchemeng.2023.108453
Maurya, M. R., Rengaswamy, R., & Venkatasubramanian, V. (2004). Application of signed digraphs-based analysis for fault diagnosis of chemical process flowsheets. Engineering Applications of Artificial Intelligence, 17(5), 501–518.
Rieth, C. A., Amsel, B. D., Tran, R., & Cook, M. B. (2017). Additional Tennessee Eastman process simulation data for anomaly detection evaluation. Harvard Dataverse, 1.
Ruff, L., Vandermeulen, R., Görnitz, N., Deecke, L., Siddiqui, S. A., Binder, A., ... Kloft, M. (2018). Deep one-class classification. In Proceedings of the International Conference on Machine Learning (ICML) (pp. 4393–4402).
Sakurada, M., & Yairi, T. (2014). Anomaly detection using autoencoders with nonlinear dimensionality reduction. In Proceedings of the MLSDA Workshop (pp. 4–11).
Tidriri, K., Chatti, N., Verron, S., & Tiplica, T. (2018). Model-based fault detection and diagnosis of complex chemical processes: A case study of the Tennessee Eastman process. Proceedings of the IMechE, Part I: Journal of Systems and Control Engineering, 232(6), 742–760. doi: 10.1177/0959651818764510
Veličković, P., Cucurull, G., Casanova, A., Romero, A., Liò, P., & Bengio, Y. (2018). Graph attention networks. Retrieved from https://arxiv.org/abs/1710.10903
Venkatasubramanian, V., Rengaswamy, R., Yin, K., & Kavuri, S. N. (2003). A review of process fault detection and diagnosis: Part I — Quantitative model-based methods. Computers & Chemical Engineering, 27(3), 293–311.
Wu, Z., Pan, S., Chen, F., Long, G., Zhang, C., & Yu, P. S. (2021). A comprehensive survey on graph neural networks. IEEE Transactions on Neural Networks and Learning Systems, 32(1), 4–24.
Zhao, H., Wang, Y., Duan, J., Huang, C., Cao, D., Tong, Y., ... Zhang, Q. (2020). Multivariate time-series anomaly detection via graph attention network. In Proceedings of the IEEE International Conference on Data Mining (ICDM) (pp. 841–850).
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
The Prognostic and Health Management Society advocates open-access to scientific data and uses a Creative Commons license for publishing and distributing any papers. A Creative Commons license does not relinquish the author’s copyright; rather it allows them to share some of their rights with any member of the public under certain conditions whilst enjoying full legal protection. By submitting an article to the International Conference of the Prognostics and Health Management Society, the authors agree to be bound by the associated terms and conditions including the following:
As the author, you retain the copyright to your Work. By submitting your Work, you are granting anybody the right to copy, distribute and transmit your Work and to adapt your Work with proper attribution under the terms of the Creative Commons Attribution 3.0 United States license. You assign rights to the Prognostics and Health Management Society to publish and disseminate your Work through electronic and print media if it is accepted for publication. A license note citing the Creative Commons Attribution 3.0 United States License as shown below needs to be placed in the footnote on the first page of the article.
First Author et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.