EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Physics-Informed Multi-Task Soft Sensor for Uncertainty-Aware Slag Tin Content Prediction in Top-Blown Tin Smelting

Zhaojun Ma, Jubo Peng, Xiaojun Zhou, Kui Xie, Zhang Zhang

Sensors · 14.08.2026 · 26 (16) · 5157

DOI: 10.3390/s26165157 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 14.08.2026. Verfasst von Zhaojun Ma und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Resources and Environment, Yunnan University, Kunming 650500, China. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China, Yunnan Fundamental Research Projects.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Slag tin content is a key indicator of reduction efficiency and metal recovery in top-blown tin smelting, but delayed and sparse laboratory assays hinder online monitoring. This study proposes a physics-informed multi-task soft sensor (PI-MTSS) combining a shared temporal convolutional encoder, self-supervised future state reconstruction, a soft monotonicity constraint based on a simplified CO–temperature reduction-potential indicator, and Monte Carlo dropout for epistemic uncertainty estimation. The model was evaluated on a continuous 30-day dataset comprising 43,200 one-minute process observations and 240 laboratory assays, including 40 chronologically held-out test labels. Across five random seeds, PI-MTSS achieved an RMSE of 0.198, an MAE of 0.145, and an R2 of 0.889. Across three rolling-origin evaluations, it obtained a mean RMSE of 0.199 and maintained the lowest RMSE among the evaluated methods. The complete model reduced the physical violation rate to 3.5%. At 95% nominal coverage, its epistemic intervals achieved a PICP of 96.5% and an NMPIW of 0.245; the ACE across six nominal levels was 0.021, and uncertainty correlated positively with absolute error (ρs=0.691). These results support the potential of PI-MTSS for minute-scale monitoring and operator decision support under sparse labels, although validation remains limited to one month from a single furnace.

Schlagwörter

Stichwörter der Autor:innen

industrial process monitoringmulti-task learningphysics-informed learningslag tin contentsoft sensorsparse labelstop-blown tin smeltinguncertainty quantification

Autor:innen und Institutionen

  • School of Resources and Environment, Yunnan University, Kunming 650500, China.
  • Yunnan Tin Group (Holding) Co., Ltd., No. 471 Minhang Road, Guandu District, Kunming 650200, China.
  • School of Automation, Central South University, Changsha 410083, China.

Förderung

  • National Natural Science Foundation of China
  • Yunnan Fundamental Research Projects

Zitieren

APA

Ma, Z., Peng, J., Zhou, X., Xie, K., & Zhang, Z. (2026). A Physics-Informed Multi-Task Soft Sensor for Uncertainty-Aware Slag Tin Content Prediction in Top-Blown Tin Smelting. Sensors, 26(16), 5157. https://doi.org/10.3390/s26165157

Vancouver

Ma Z, Peng J, Zhou X, Xie K, Zhang Z. A Physics-Informed Multi-Task Soft Sensor for Uncertainty-Aware Slag Tin Content Prediction in Top-Blown Tin Smelting. Sensors. 2026;26(16):5157. doi:10.3390/s26165157

BibTeX
@article{Ma2026,
  author = {Ma, Zhaojun and Peng, Jubo and Zhou, Xiaojun and Xie, Kui and Zhang, Zhang},
  title = {A Physics-Informed Multi-Task Soft Sensor for Uncertainty-Aware Slag Tin Content Prediction in Top-Blown Tin Smelting},
  journal = {Sensors},
  year = {2026},
  volume = {26},
  number = {16},
  pages = {5157},
  doi = {10.3390/s26165157}
}
Kostenlos starten →