Simplified State Space Building Energy Model and Transfer Learning Based Occupancy Estimation for HVAC Optimal Control

An energy model for non-residential buildings based on state space equations has been developed and experimentally validated. This model can predict with precision the Heating Ventilation and Air Conditioning (HVAC) system energy consumption considering several variables including temperature setpoints, external temperature, humidity, wind speed, solar irradiance, and occupancy. This last variable is estimated through deep transfer learning (TL). The proposed simplified energy model can be easily exploited by Model Predictive Control (MPC) strategies for HVAC optimal control.