Optimizing energy using predictive control of opening windows behavior

Energy consumption in buildings is a salient factor in environmental sustainability and economic efficiency. Heating, ventilation, and air conditioning (HVAC) systems frequently account for a substantial proportion of energy usage, resulting in elevated operational expenditures and resource exhaustion. This study proposes a predictive control framework that utilizes fuzzy logic and the Random Forest algorithm to optimize energy consumption by predicting window-opening behavior. The model dynamically adjusts HVAC operations based on environmental conditions, occupancy, and hourly energy demand, achieving energy savings of up to 44% compared to another predictive controller based on real data of occupancy. The efficacy of the proposed approach is substantiated by simulation results, which underscore its capacity to ensure thermal comfort while concurrently enhancing energy efficiency.