Energy- and process real-time optimization through hybrid modeling - a case from Viking Malt A/S

The production of barley malt is an energy-intensive process due to the need for cooling, drying, and heating. This work introduces model-based real-time energy- and process optimization for an industrial malting process. A hybrid model is introduced for the germination stage in the malting process by combining population- and mass balances with probabilistic ML, providing accurate predictions with no indication of overfitting when applied to historical process data. Based on the model predictions, optimal setpoint trajectory for the temperature of process air passing through the grain bed and the water/gibberellic acid addition are recommended for operators of the malting plant. Both the model and optimization algorithm are deployed to servers at the malting plant, and the recommendations are presented in a user-friendly dashboard at a dedicated monitor in the control room, enabling the decision-making in the process to be objective-driven rather than human-driven. No previous work has been found in the literature on applying hybrid model-based real-time optimization to the germination stage in an industrial-scale malting process.

Recommended citation: Jul-Rasmussen, P., Aouichaoui, A. R., Korzepa, M., Engelbrecht, K., Kongsgaard, K., & Huusom, J. K. (2024). Energy-and process real-time optimization through hybrid modeling-a case from Viking Malt A/S. In Computer Aided Chemical Engineering (Vol. 53, pp. 1645-1650). Elsevier. https://doi.org/10.1016/B978-0-443-28824-1.50275-1