Physics-Guided Soft Sensing under Sparse Laboratory Measurements and Temporal Regime Shifts in Industrial Monoammonium Phosphate Production
4 affiliations
- 1Moscow Institute of Physics and Technology (National Research University
- 2Molecular Electronics Research Institute
- 3JSC (MERI
- 4UralAiTech LLC
Abstract
Industrial plants collect high-frequency telemetry while critical quality variables may be measured only every few hours. We study soft sensing for an industrial monoammonium phosphate process using 260,642 one-minute records and 1,325 paired laboratory observations. The grey-box model reconstructs latent acid properties from conductivity and combines them with material-balance calculations to estimate molar ratio and downstream density. Validation is entirely chronological: nine pre-gap rolling tests, a short-term regime-transfer test, a strict external post-gap test on 438 future laboratory pairs, and ten post-gap rolling comparisons. To prevent uncontrolled extrapolation, the conductivity input to the fitted latent-property mappings is limited to the range observed in the corresponding training data. Ridge and LSBoost remain more accurate than the physics-guided model, particularly in the independent long-horizon evaluation. The results show that physical structure improves interpretability but does not by itself ensure temporal robustness, motivating regime-aware monitoring, recalibration, and fallback logic for industrial virtual telemetry.
Related
- Laboratory-to-Industrial Translation in Plant-Derived Bioproduct Valorization: A Critical Integrative Review and Proposed Framework for Route-Level Evidence Assessment
- Temporally Validated and Laboratory-Reconciled Soft Sensing of Xylene-Soluble Fraction in Industrial Polypropylene Manufacturing
- In Situ FTIR Spectroscopy as a Kinetic Study Tool for Potassium Butyl Xanthate Industrial Production
- Versatile and efficient synthesis of bis(phosphate) alkyl esters and first evidence for their antibacterial activity against Gram-positive bacteria
- Opposite Effects of Removing Ethylene Carbonate from the Electrolyte on Ni-Rich Layered Oxide and Mn-Rich Olivine Phosphate Cathodes
BibTeX
@misc{karnup2026physicsguided,
title = {{Physics-Guided Soft Sensing under Sparse Laboratory Measurements and Temporal Regime Shifts in Industrial Monoammonium Phosphate Production}},
author = {German A. Karnup and Andrey N. Sychev and Oleg A. Telminov and Evgeny S. Gornev},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15009979/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15009979/v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Physics-Guided Soft Sensing under Sparse Laboratory Measurements and Temporal Regime Shifts in Industrial Monoammonium Phosphate Production
AU - German A. Karnup
AU - Andrey N. Sychev
AU - Oleg A. Telminov
AU - Evgeny S. Gornev
PY - 2026
DA - 2026/10/06
PB - ChemRxiv
DO - 10.26434/chemrxiv.15009979/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15009979/v1
AB - Industrial plants collect high-frequency telemetry while critical quality variables may be measured only every few hours. We study soft sensing for an industrial monoammonium phosphate process using 260,642 one-minute records and 1,325 paired laboratory observations. The grey-box model reconstructs latent acid properties from conductivity and combines them with material-balance calculations to estimate molar ratio and downstream density. Validation is entirely chronological: nine pre-gap rolling tests, a short-term regime-transfer test, a strict external post-gap test on 438 future laboratory pairs, and ten post-gap rolling comparisons. To prevent uncontrolled extrapolation, the conductivity input to the fitted latent-property mappings is limited to the range observed in the corresponding training data. Ridge and LSBoost remain more accurate than the physics-guided model, particularly in the independent long-horizon evaluation. The results show that physical structure improves interpretability but does not by itself ensure temporal robustness, motivating regime-aware monitoring, recalibration, and fallback logic for industrial virtual telemetry.
ER -