HVACLearn: A reinforcement learning based occupant-centric control for thermostat set-points

摘要: In this paper, we present a Reinforcement Learning (RL) based Occupant-Centric Controller (OCC) for thermostats, HVACLearn. Monitoring indoor air temperature, occupancy, and thermal vote, the agent learns unique occupant behavior environments calculates adaptive thermostat set-points to balance between comfort energy efficiency. We simulated HVACLearn performance in single office with models from literature (i.e., occupancy vote). Compared reference controller, reduced number of button presses (too hot) significantly, while consuming same or less cooling energy. For heating, resulted almost cold) slightly heating consumption.