Efficient Data-Driven MPC for Demand Response of Commercial Buildings

TL;DR

The paper tackles energy-efficient demand response for small commercial buildings by marrying an Input Convex Recurrent Neural Network (ICRNN) with a mixed-integer convex MPC. This framework enables real-time control with discrete rooftop-unit actuators and a data-driven, tractable optimization that guarantees global optimality under relaxation while supporting a real-time DR bidding strategy. Through numerical studies on a two-zone Modelica-based building, the convex ICRNN-MPC achieves lower energy use, reduced peak demand, and substantially improved thermal comfort across flat-rate, TOU, and DR programs, outperforming linear, nonconvex, and greedy baselines. The work demonstrates practical viability for market participation and paves the way for scalable aggregation and distributed optimization in building-level demand response.

Abstract

Model predictive control (MPC) has been shown to significantly improve the energy efficiency of buildings while maintaining thermal comfort. Data-driven approaches based on neural networks have been proposed to facilitate system modelling. However, such approaches are generally nonconvex and result in computationally intractable optimization problems. In this work, we design a readily implementable energy management method for small commercial buildings. We then leverage our approach to formulate a real-time demand bidding strategy. We propose a data-driven and mixed-integer convex MPC which is solved via derivative-free optimization given a limited computational time of 5 minutes to respect operational constraints. We consider rooftop unit heating, ventilation, and air conditioning systems with discrete controls to accurately model the operation of most commercial buildings. Our approach uses an input convex recurrent neural network to model the thermal dynamics. We apply our approach in several demand response (DR) settings, including a demand bidding, a time-of-use, and a critical peak rebate program. Controller performance is evaluated on a state-of-the-art building simulation. The proposed approach improves thermal comfort while reducing energy consumption and cost through DR participation, when compared to other data-driven approaches or a set-point controller.

Figures (5)

• Figure 1: Architecture of an .
• Figure 2: Thermal dynamics fitting comparison.
• Figure 3: Controller comparison: profiles during 2 simulated days under a flat-rate pricing.
• Figure 4: Power profiles for the demand bidding program.
• Figure 5: Comparison of objective function values based on computation time.