Self-Constrained Weighted Garrote Ordered Memory Network With Adaptive Distribution Learning for Industrial Soft Sensor
IEEE Transactions on Automation Science and Engineering · 2026 · 23 · 6537-6549
DOI: 10.1109/tase.2026.3671808 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Automation Science and Engineering, 2026. Verfasst von Lin Sui und 3 weiteren Autor:innen. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 40 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Basic Research Program of Jiangsu, Fundamental Research Funds for the Central Universities.
Abstract (Originalsprache)
Chemical processes involve complex operational mechanisms and high-dimensional variables with multiscale dependencies and temporal covariate shifts, posing significant challenges for soft sensing implementation. To address these issues, a self-constrained weighted garrote ordered memory network with adaptive distribution learning is proposed for industrial soft sensor modeling. Specifically, an improved long short-term ordered memory (LSTOM) architecture with a composite hierarchical information update mechanism is designed to capture multiscale dynamic features from industrial time-series data. Meanwhile, an adaptive distribution learning module is embedded into the LSTOM architecture to match high-order feature distributions across different temporal periods via maximally dissimilar segments. With the composite hierarchical update mechanism and embedded adaptive distribution learning, the proposed approach improves generalization in soft sensor modeling under multi-timescale dependencies and temporal covariate shifts. Finally, the shrinkage coefficients of the self-constrained weighted garrote are embedded into the LSTOM’s input weights to eliminate redundant variables and promote structural sparsity. The effectiveness of the proposed approach is validated on a public industrial-scale penicillin fermentation process and a real-world flue gas desulfurization case. Note to Practitioners—Soft sensors are widely used in industrial processes to estimate key variables that are difficult or costly to measure directly. However, challenges such as high-dimensional input data, multiscale temporal dependencies, and distribution shifts over time often degrade their performance. This work presents a practical modeling framework that enhances soft sensor reliability and interpretability by combining adaptive distribution learning with a sparsity-promoting ordered memory network. The proposed method automatically eliminates redundant input variables and captures complex temporal patterns in time-series data under temporal covariate shifts, making it suitable for deployment in chemical and energy industries. Moreover, the model’s variable selection results were reviewed against known process mechanisms and expert knowledge, ensuring that its outputs and reasoning are consistent with practical domain understanding. Its effectiveness has been validated on both a public penicillin fermentation dataset and a real-world flue gas desulfurization process, demonstrating its potential for improving process monitoring and control in dynamic industrial environments.
Förderung
- National Natural Science Foundation of China
- Basic Research Program of Jiangsu
- Fundamental Research Funds for the Central Universities
- National Foreign Expert Project
Zitieren
APA
Sui, L., Shi, X., Holderbaum, W., & Xiong, W. (2026). Self-Constrained Weighted Garrote Ordered Memory Network With Adaptive Distribution Learning for Industrial Soft Sensor. IEEE Transactions on Automation Science and Engineering, 23, 6537-6549. https://doi.org/10.1109/tase.2026.3671808
Vancouver
Sui L, Shi X, Holderbaum W, Xiong W. Self-Constrained Weighted Garrote Ordered Memory Network With Adaptive Distribution Learning for Industrial Soft Sensor. IEEE Transactions on Automation Science and Engineering. 2026;23:6537-6549. doi:10.1109/tase.2026.3671808
BibTeX
@article{Sui2026,
author = {Sui, Lin and Shi, Xudong and Holderbaum, William and Xiong, Weili},
title = {Self-Constrained Weighted Garrote Ordered Memory Network With Adaptive Distribution Learning for Industrial Soft Sensor},
journal = {IEEE Transactions on Automation Science and Engineering},
year = {2026},
volume = {23},
pages = {6537--6549},
doi = {10.1109/tase.2026.3671808}
}
EUVIMED