Dynamic Mechanism Design for Human-In-The-Loop Control of Building Energy Consumption

Maximilian Schütte1, Ashish Hota2, Annika Eichler1, John Lygeros1

  • 1ETH Zurich
  • 2Indian Institute of Technology (IIT), Kharagpur

Details

11:20 - 11:40 | Wed 10 Jul | Franklin 7 | WeA07.5

Session: Control & Energy Management of Building Systems

Abstract

Integrating the preferences of building occupants for efficient building energy management has the potential for significant energy savings. In this paper, we present a new dynamic mechanism that achieves ex-post incentive compatibility, i.e., the occupants reveal their privately-held preferences truthfully in every time period. These preferences are then incorporated in a receding horizon control scheme to jointly minimize the cost of energy purchase and the discomfort experienced by the occupants. We evaluate the performance of our scheme with a baseline heating policy based on standardized thermal comfort bounds, and the classical dynamic pivot mechanism via extensive numerical simulations of a sample building. We illustrate that due to the integration of individual preferences, both occupant discomfort and energy consumption can be greatly reduced compared to the baseline heating policy. Furthermore, while the dynamic pivot mechanism requires occupants to pay for their comfort, our payment scheme rewards the occupants whose participation leads to energy savings, making their participation in the scheme more palatable.