EUVIMEDEUVIMED entdecken
Crossref

Enhancing Quality Prediction in Industrial Processes With Spatiotemporal Attention-Based Data Techniques

Li Zhu, Xiaoyu Wang, Zhe Li, Junghui Chen

IEEE Transactions on Automation Science and Engineering · 2026 · 23 · 9107-9118

DOI: 10.1109/tase.2026.3682860 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Automation Science and Engineering, 2026. Verfasst von Li Zhu und 3 weiteren Autor:innen. Die Arbeit nennt 28 Literaturangaben. Gefördert von: Liaoning Provincial Natural Science Foundation of China, National Science and Technology Council, Taiwan,.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The intricate and dynamic characteristics of time series data in industrial processes necessitates sophisticated data-driven soft sensor methodologies for modeling. However, the performance of such models often heavily hinges data quality, leading to potential neglect of critical spatial and temporal correlations essential for comprehending model behavior. To address these challenges, this study proposes a novel approach integrating spatiotemporal-based graph convolutional networks with attention-based preprocessing for soft sensor modeling. This framework not only enhances model reliability by extracting high-quality time series data but also captures localized spatiotemporal correlations vital for comprehending the intricate interactions among variables in the soft sensor. Additionally, the integration of attention-based preprocessing enables adaptive updates to the soft sensor model, thereby ensuring accurate prediction of process quality. The effectiveness and applicability of the proposed method are demonstrated through experimentation on both simulated processes and real-world applications in the ethylene oxide industry. Importantly, this data-driven approach to constructing transfer entropy aligns with fundamental a priori knowledge, enhancing the interpretability of the model results. Note to Practitioners—The dynamic and complex nature of industrial time series data requires advanced soft sensor methodologies. However, conventional models are highly sensitive to raw data quality, often missing critical spatiotemporal correlations. To address this, we propose a novel framework integrating spatiotemporal graph convolutional networks with attention-based data preprocessing. This approach enhances reliability by filtering high-quality features and capturing localized spatiotemporal interactions among process variables. Furthermore, the attention mechanism enables adaptive model updates, ensuring accurate process quality predictions. Validated on both simulated and real-world ethylene oxide processes, our data-driven use of transfer entropy aligns with fundamental a priori knowledge, significantly improving model interpretability for practical industrial deployment.

Förderung

  • Liaoning Provincial Natural Science Foundation of China
  • National Science and Technology Council, Taiwan,

Zitieren

APA

Zhu, L., Wang, X., Li, Z., & Chen, J. (2026). Enhancing Quality Prediction in Industrial Processes With Spatiotemporal Attention-Based Data Techniques. IEEE Transactions on Automation Science and Engineering, 23, 9107-9118. https://doi.org/10.1109/tase.2026.3682860

Vancouver

Zhu L, Wang X, Li Z, Chen J. Enhancing Quality Prediction in Industrial Processes With Spatiotemporal Attention-Based Data Techniques. IEEE Transactions on Automation Science and Engineering. 2026;23:9107-9118. doi:10.1109/tase.2026.3682860

BibTeX
@article{Zhu2026,
  author = {Zhu, Li and Wang, Xiaoyu and Li, Zhe and Chen, Junghui},
  title = {Enhancing Quality Prediction in Industrial Processes With Spatiotemporal Attention-Based Data Techniques},
  journal = {IEEE Transactions on Automation Science and Engineering},
  year = {2026},
  volume = {23},
  pages = {9107--9118},
  doi = {10.1109/tase.2026.3682860}
}
Kostenlos starten →