Building Energy Reinforcement Strategy based on Data-Driven Methods

The construction industry has become a significant contributor to global energy consumption and greenhouse gas emissions. The inherent volatility in energy consumption data and the time-intensive nature of traditional certification processes pose substantial challenges for Building Energy Management Systems. This study presents an innovative AI-driven framework integrating Particle Swarm Optimization and enumeration techniques to address these dual challenges. Utilizing a Machine Learning approach, with the Random Forest (RF) model achieving superior predictive accuracy (R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> = 0.9644), we developed three tailored strategies to mitigate energy consumption irregularities. These strategies emphasize carbon reduction, economic viability, and operational efficiency, specifically designed for buildings with varying solar radiation exposure, photovoltaic absorption ratios, and floor areas. Key outcomes include a PSO-based model that delivers 21.2% energy savings, customized strategies yielding a Net Present Value of £4,619.94 with a payback period of 3.25 years, and a comprehensive multidimensional evaluation framework that harmonizes technical, economic, and sustainability performance metrics. This work establishes a methodological innovation for optimizing energy management across diverse building typologies.