A proposed hybrid machine learning and physics-based optimization framework for energy-efficient smart buildings

Abstract

Due to the complex and dynamic structure of building energy systems, improving energy efficiency in smart buildings while ensuring occupant comfort has proven difficult. This paper proposes a hybrid framework which combines machine learning, physics-based thermal model of the building and optimization for intelligent building energy management. The proposed framework does not just use data-driven or physics-based approaches but rather integrates the complementary benefits to enhance energy prediction and control performance. The dataset for ASHRAE Great Energy Predictor III was used for evaluating the framework. After data preprocessing, the data was randomly split into three sets for training (70%), validation (15%) and testing (15%). The model performance was evaluated by the root mean square error (RMSE), mean absolute error (MAE) and energy savings, and compared to the baseline models of Linear Regression, Random Forest and Artificial Neural Network. Compared to the comparison methods on the benchmark dataset, the proposed system achieved the best results with RMSE 0.24 kW, MAE 0.19 kW, and up to 26.4% of energy savings when evaluated on the complete ASHRAE Great Energy Predictor III benchmark dataset. This work shows the effectiveness of incorporating machine learning with physics-based thermal modelling and optimization in enhancing prediction accuracy and operational energy efficiency. Overall, the proposed hybrid framework significantly improves HVAC energy prediction accuracy and energy efficiency compared with conventional methods, demonstrating its potential as a practical intelligent energy management solution for smart buildings.