A Quantum Annealing-Based Three-Stage Scheduling Strategy for Multi-Stack Fuel Cell Hybrid Power Systems

Fuel cell hybrid power systems (FCHPS) face significant challenges due to the non-convex nature of their optimization problems, especially in high-power applications with multi-stack configurations that involve numerous start-stop decisions, introducing a high number of binary variables. To address these issues, this paper presents a quantum annealing (QA)-based three-stage scheduling strategy for multi-stack solid oxide fuel cell (SOFC)-based fuel cell hybrid power systems (FCHPS). The proposed method decouples the decision-making process across different timescales-day-ahead, intra-day, and real-time-tailoring decisions to the dynamics of various power sources within the FCHPS. In the day-ahead stage, global predictions inform the startup and shutdown of SOFCs; in the intra-day stage, short-term predictions refine power outputs; and in the real-time stage, adjustments are made to respond to immediate operational conditions. Quantum annealing is introduced to expedite the solution of the large-scale, binary optimization problems inherent in multi-stack configurations. A OPAL-RT-based experimental platform is used to validate the proposed strategy. In addition, a comparison between the proposed method and conventional methods is conducted, indicating that the proposed QA-based approach significantly speeds up the computation process-being 49.89 times faster than the dual model (DMPC) predictive control method and 22.25 times faster than the Gurobi-based method. It also optimizes the overall operational cost, achieving a reduction in the total objective function value by approximately 10.62% compared to the Gurobi-based method, and by 14.66% compared to the DMPC method.