Flexible operation of virtual power plant enabled integrated electricity-heating system under multiple uncertainties via distributionally robust model predictive control

Virtual power plant (VPP) can incorporate various electric infrastructures, e.g., data centers (DCs) and electric vehicles (EVs), creating multiple uncertainties and challenges for the operation of integrated electricity–heating system (IEHS). This paper focuses on the flexible operation problem of VPP-enabled IEHS under both static and dynamic uncertainties. First, the temporal shifting flexibility of workloads from DCs is modeled. Second, a novel metric-based distributionally robust model predictive control (DRMPC) framework is introduced to address both static uncertainties from renewable energy and dynamic uncertainties from EV charging behaviors. Third, the dynamic uncertainties are reformulated as ambiguity tubes, and distributionally robust bounds for both dynamic and static uncertainties are determined using DRMPC. Through ambiguity tubes and distributionally robust optimization, the stochastic MPC system is converted into a nominal one. Case studies validate the effectiveness of the proposed approach.