Joint Optimization of Generation and Maintenance for Large Cascaded Hydropower With Complex Operational Requirements

The generation and maintenance of cascade hydropower face strong spatiotemporal coupling under multiple power grids’ flexible regulation demands, manifested by capacity competition, time mismatch, and nonlinear constraint interactions. This article proposes a joint optimization model for generation and maintenance, based on China’s largest mainstream cascade hydropower system. A mapping between daily average and peak generation is established to identify feasible maintenance windows, enable long-term daily scheduling, and meet hourly load-following requirements. This mapping is incorporated into a 730-days panoramic model which considers critical constraints including cross-year maintenance continuity, seasonal energy allocation, monthly generation trends, maintenance restrictions during peak supply periods, and transmission limits. The model’s complex logic constraints are reformulated by the Big-M method, becoming a tractable mixed-integer linear programming formulation. Case studies show that the proposed approach substantially alleviates current coordination difficulties, increases hydropower generation by 36 million kWh, and eliminates downtime during peak-demand hours. Sensitivity analysis further reveals that tighter peak-shaving requirements significantly reduce maintenance flexibility and increase reserve pressure.