In this section
@article{Pessina:2026:10.1016/j.dche.2026.100292,
author = {Pessina, D and Tian, T and Watson, O and Heng, J and Papathanasiou, M},
doi = {10.1016/j.dche.2026.100292},
journal = {Digital Chemical Engineering},
title = {Transfer learning of data-driven crystallisation processes via constrained neural ordinary differential equations},
url = {http://dx.doi.org/10.1016/j.dche.2026.100292},
volume = {18},
year = {2026}
}
TY - JOUR
AB - 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 andsparse 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-dataregimes.
AU - Pessina,D
AU - Tian,T
AU - Watson,O
AU - Heng,J
AU - Papathanasiou,M
DO - 10.1016/j.dche.2026.100292
PY - 2026///
SN - 2772-5081
TI - Transfer learning of data-driven crystallisation processes via constrained neural ordinary differential equations
T2 - Digital Chemical Engineering
UR - http://dx.doi.org/10.1016/j.dche.2026.100292
VL - 18
ER -