Publications

You can also find my articles on my Google Scholar profile.

Transfer learning for cross-building forecasting of building energy and indoor air temperature in model predictive control applications

Published in Journal of Building Engineering, 2025

Download the paper here When applying Model Predictive Control (MPC) for Heating, Ventilation and Air Conditioning (HVAC) systems in buildings, accurate forecasting of short-term energy demand and indoor air condition profiles is essential. However, new or retrofitted buildings lack sufficient operation data to develop precise data-driven models. This study investigates transfer learning techniques to enhance the forecasting performance of black-box models under limited data conditions. Specifically, we leverage synthetic data from an open-source EnergyPlus building model to pre-train three neural network models, which are then transferred to a real building and fine-tuned with limited measurements. The results indicate that incorporating synthetic data into the pre-training phase significantly enhances the forecasting accuracy for building and HVAC energy, as well as indoor air temperature profiles, over a 12-h horizon with 15-min intervals. The study underscores the potential of combining transfer learning with synthetic data to address data limitations, extending the applicability of learning-based MPC in real-world buildings. Recommended citation: Dou, H., Zhang, K., 2025. Transfer learning for cross-building forecasting of building energy and indoor air temperature in model predictive control applications. Journal of Building Engineering 111, 113341. https://doi.org/10.1016/j.jobe.2025.113341 Read more

From theory to practice: A critical review of model predictive control field implementations in the built environment

Published in Applied Energy, 2025

Download the paper here While the potential of model-based predictive control (MPC) to improve building operation is widely acknowledged, its implementation has not yet become a mainstream practice in the building operation industry. This review paper explores the scientific literature documenting MPC field implementations in actual buildings. The goal is twofold: (a) to identify critical features in the deployment of MPC strategies, including the targeted building types and applications, systems controlled, expected benefits, software used, as well as common issues encountered (and successful measures to overcome these issues); (b) to evaluate the benefits of MPC based on the reported information from real-life implementations. Aspects analyzed include drivers and energy contexts, control-oriented modelling approaches, optimization routines, and performance evaluation methods. Results show that most practical studies focussed on buildings with a floor area under 10,000 m2, often even less than 1000 m2. MPC applications were varied, ranging from setpoint tracking and building conditioning to the optimization of the operation of thermal energy storage and photovoltaic panels and/or battery systems. MPC consistently yields significant benefits, with average savings of 30 % for thermal energy, 25 % for electricity use, 25 % for energy costs, 26 % for peak power and 17 % for GHG emissions, obtained under an average field-testing duration of 41 days. Recommended citation: Saloux, E., Candanedo, J.A., Vallianos, C., Morovat, N., Zhang, K., 2025. From theory to practice: A critical review of model predictive control field implementations in the built environment. Applied Energy 393, 126091. https://doi.org/10.1016/j.apenergy.2025.126091 Read more