A novel occupant-centric control framework for building to urban scale hybrid-energy simulation

Decarbonizing urban energy systems requires effective reduction strategies. While AI/ML models provide insights, simulation-based frameworks offer reliable long-term assessments. A key challenge is capturing interactions between indoor environments, occupants, and HVAC usage.

This study introduces aPKf, a hybrid occupancy model integrating prior space usage knowledge with real-time probabilistic transitions. Unlike traditional models, aPKf adapts based on past occupancy states, improving accuracy using real-time energy and air quality data. Simulations of 51 office buildings show aPKf predicts a 23.3–24.4% rise in total energy demand and 37.6–39.6% in cooling demand, underscoring the importance of behavior-driven modeling in post-COVID cities.