Optimal scheduling of building energy system with integrated virtual energy storage based on multi-task model predictive control

The development of building energy system (BES) integrating solar photovoltaic (PV) can greatly reduce the electricity cost and require more intelligent scheduling methods. The virtual energy storage (VES) is an innovative, economical and efficient technology that gives building energy storage capability using the thermal inertia characteristics and provides more flexibility for the optimal scheduling scheme of BES. This paper proposes an optimal scheduling method for BES integrating VES based on multi-task model predictive control (MPC). First, considering the heat transfer theory of the building envelope, the VES model with three parameters virtual charge and discharge power (VCDP), virtual capacity (VC), and virtual state of charge (VSOC) is established to quantify the storage capacity of the VES. Then, a state space model of the BES with integrated VES is developed, and an optimal scheduling scheme is designed based on multi-task MPC to simultaneously guarantee occupant thermal comfort, energy cost and PV utilization. Finally, a residential building developed on simulation software is used as validation, and the results show that the cost of multi-task MPC scheduling approach can improve by 37.40 %, 4.02 % and 46.50 % respectively compared with other methods, which indicates that the proposed method has better economic performance.