Region-Driven Two-Layer Refined Scheduling for Multi-Stack-Integrated Alkaline Electrolyzer in Wind-Hydrogen System

To mitigate the stability impacts of renewable intermittency on power systems, on-site accommodation via Power-to-Hydrogen (P2H) stands out as a promising solution. As the most mature P2H technology, balance-of-plant (BOP)-shared multi-stack integrated alkaline water electrolyzer (MSIAWE) systems offer both high flexibility and cost advantages. However, accommodating highly fluctuating wind power requires scheduling MSI-AWE system at minute-level intervals, posing significant computational burdens for real-time energy management. To address this, this paper first introduces the concept of operational regions to efficiently characterize the time-varying power boundaries of MSI-AWE systems, along with a pruning-based algorithm to accelerate region computation. Further, a minute-level model predictive control (MPC)-based wind-hydrogen operational framework is proposed, which decouples the refined scheduling problem into region-embedded upper-layer power allocation and lower-layer electrolyte/cooling flow regulation. Case studies demonstrate that the two-layer framework extends full wind power utilization duration by 8.7%, improves state regulation accuracy by 4.6%, and reduces average computation time by 50.2% compared to traditional single-layer approaches, validating its effectiveness in enhancing renewable energy accommodation and facilitating smart grid integration.