Stochastic model predictive control for the optimal operation of office buildings

This paper developed a novel stochastic model predictive control (SMPC) strategy to enhance the operational efficiency of office buildings. Firstly, an improved state space model encompassing temperature and relative humidity simultaneously is developed to accurately characterize the thermal comfort condition within the office building. Then, given the obtained comprehensive model, a new SMPC approach is proposed based on chance constraints to minimize energy consumption while guaranteeing thermal comfort for occupants. Besides, the feasibility and stability properties of the SMPC are demonstrated theoretically. Finally, the proposed SMPC method is verified through a real office building located in Xi'an, China, and the result shows that compared to the conventional ONOFF and MPC control strategies, the SMPC can achieve 39.1 % and 33.3 % energy-saving and less temperature and relative humidity requirement violations.