A hybrid approach of transfer learning and physics-informed modelling: Improving dissolved oxygen concentration prediction in an industrial wastewater treatment plant

• Published in: Chemical engineering science 2025-02, Vol.304, p.121088, Article 121088
• Main Authors: Koksal, Ece Serenat, Aydin, Erdal
• Format: Article
• Language: English
• Subjects: Long short-term memory; Physics-informed transfer learning; Recurrent neural networks; Transfer learning; Wastewater treatment unit
• ISSN: 0009-2509
• DOI: 10.1016/j.ces.2024.121088

•Transfer learning predicts dissolved oxygen concentration in an industrial wastewater treatment plant.•Knowledge transfer from open-source model improves predictions despite system dissimilarities.•Knowledge transfer from a similar plant improves predictions despite limited and noisy data.•Physics informed transfer learning boosts prediction performance significantly. Constructing first principles models is a challenging task for nonlinear and complex systems such as a wastewater treatment unit. In recent years, data-driven models are widely used to overcome the complexity. However, they often suffer from issues such as missing, low quality or noisy data. Transfer learning is a solution for this issue where knowledge from another task is transferred to target one to increase the prediction performance. In this work, the objective is increasing the prediction performance of an industrial wastewater treatment plant by transferring the knowledge of (i) an open-source simulation model, capturing process physics, albeit with dissimilarities to the target plant, (ii) another industrial plant characterized by noisy and limited data but located in the same refinery, and (iii) constructing a physics informed transfer learning model by combining (i) and (ii). The results demonstrated that test and validation performance are improved up to 27% and 59%, respectively.