Transfer learning of data-driven crystallisation processes via constrained neural ordinary differential equations
File(s)
Published version
Author(s)
Pessina, Daniele
Tian, Tony
Watson, Oliver
Heng, Jerry
Papathanasiou, Maria
Type
Journal Article
Abstract
Modelling complex crystallisation processes remains challenging due to limited ex perimental datasets, high measurement noise, and the need for generalisability across varying operating conditions. Neural Ordinary Differential Equations (NODEs) and transfer learning (TL) offer promising tools to overcome these limitations by provid ing data-efficient, flexible, and transferable modelling frameworks. This work investigates the use of NODEs to model protein crystallisation dynamics under data-scarce conditions. A NODE trained on a data-rich source system successfully captures solute consumption and particle size dynamics, but when applied to data-sparse target sys tems, scratch-trained NODEs exhibit limited generalisation and unphysical behaviours. To address this, several TL strategies are evaluated, including layer freezing, parame ter deviation penalisation, and system-embedding within the neural architecture. Results show that layer freezing and deviation penalty consistently improve knowledge transfer, while system-embedding offers robustness in noisy or undersampled datasets. In addition, physics-informed NODEs, constrained to enforce monotonic concentra tion decay and crystal growth, demonstrate greater stability under high noise and
sparse measurement regimes, ensuring physically consistent predictions. Overall, the combination of constrained NODEs with appropriate TL strategies provides a robust framework for accurate, transferable modelling of crystallisation systems in low-data
regimes.
sparse measurement regimes, ensuring physically consistent predictions. Overall, the combination of constrained NODEs with appropriate TL strategies provides a robust framework for accurate, transferable modelling of crystallisation systems in low-data
regimes.
Date Issued
2026-03-01
Date Acceptance
2026-02-09
Citation
Digital Chemical Engineering, 2026, 18
ISSN
2772-5081
Publisher
Elsevier
Journal / Book Title
Digital Chemical Engineering
Volume
18
Copyright Statement
© 2026 The Authors. Published by Elsevier Ltd on behalf of Institution of Chemical Engineers (IChemE). This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
License URL
Identifier
10.1016/j.dche.2026.100292
Publication Status
Published
Article Number
ARTN 100292
Date Publish Online
2026-02-16