A segmented optimization HUC model considering refined vibration zone characteristics and predictive risk control strategy

The growing penetration of renewable energy sources has compelled hydropower units to operate more frequently within or near their vibration zones, exacerbating mechanical stress and compromising system reliability. However, existing scheduling models often neglect unit-level vibration zone heterogeneity and lack predictive risk control, thereby limiting their robustness under uncertainty. To address these challenges, this study presents a segmented hydropower unit commitment (HUC) framework that synergistically integrates refined vibration zone characterization with predictive risk control. Methodologically, a risk function is constructed based on vibration zone characteristics and embedded into an enhanced ESformer model to forecast unit-level operational risks. These forecasts are incorporated into a closed-loop forecast–optimize–reforecast architecture, enabling adaptive, risk-aware scheduling. In which a dual-layer risk constraint structure—comprising a global cumulative threshold and a localized propagation limit inspired by heat conduction theory, is introduced to suppress temporal risk diffusion and maintain system stability. The model is solved via a two-layer nested strategy, where the outer layer determines unit commitment and the inner layer solves load allocation. Case studies conducted on the Ertan Hydropower Station show that the proposed method reduces operational risk by 68.7 %, conserves water by 3.5 %, and shortens cumulative unit runtime by 21.7 %. These results demonstrate that the framework enables a paradigm shift from passive constraint enforcement to proactive risk governance, offering a technically robust and scalable scheduling solution for hydropower systems under high-renewable scenarios.