针对复杂洪水地区组成下三峡与清江梯级水库协同防洪问题,选取三峡与清江流域的不同洪水地区组成作为调度模型输入,以上游防洪安全和下游防洪安全为目标,建立了三峡和清江梯级水库群多目标联合防洪优化调度模型,并采用NSGA-II算法进行求解。研究结果表明:1969年、1981年、1998年典型年设计洪水调度结果显示,相较三峡大洪水和清江大洪水情景,流域性大洪水情景下清江防洪库容投入时机更早;与实际调度方案相比,2020典型年的协调调度方案中,三峡调洪高水位降低了5.0 m,超额洪量减少了56.2亿m³,枝城、城陵矶的洪峰流量分别提高了2 603、1 035 m³/s,三峡水库末水位降低了4.2 m,协同调度的防洪效益显著。

[Objective] This study focuses on the coordinated flood control scheduling of the Three Gorges and Qingjiang cascade reservoirs under complex flood regional compositions. By selecting combinations of different flood areas in the Three Gorges and Qingjiang river basins as model inputs, this study aims to establish a multi-objective joint flood control optimal scheduling model, with innovative optimization of the activation timing and utilization methods for the reservoir group’s flood control capacity. The objectives are to fully exploit the potential of joint flood control scheduling of reservoir groups and leverage their synergistic effects in flood mitigation and disaster reduction while ensuring upstream and downstream flood control safety. [Methods] The study integrated the characteristics of flood events and regional compositions, using the frequency and intensity of extreme flood events to select combinations of flood areas as model inputs. A multi-objective joint scheduling model for the Three Gorges and Qingjiang cascade reservoirs was established, with the objective of ensuring upstream and downstream flood control safety. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) was employed to solve the model, enabling the exploration of optimal activation strategies for flood control capacity under different flood regional compositions. The model was validated using floods from typical years (1969, 1981, 1998, and 2020). The results were compared with actual scheduling schemes to highlight the benefits of coordinated scheduling. [Results] The coordinated flood control optimal scheduling model provided a broad and evenly distributed Pareto frontier, revealing a significant trade-off between the objectives of minimizing excess flood volume at control points and lowering the peak water level for flood regulation at the Three Gorges. For the typical years, the optimal peak water levels for flood regulation and excess flood volumes were as follows: 156.2-168.8 m and 13.97-25.15 billion m³ (1969), 163.3-171.0 m and 21.56-31.47 billion m³ (1981), and 161.8-171.0 m and 25.01-32.91 billion m³ (1998). The activation timing of Qingjiang flood control capacity advanced with increasing river basin flood volume, emphasizing the importance of early activation during extreme floods. Compared to the actual operation scheme, the coordinated operation scheme for the 2020 typical year reduced the maximum flood regulation water level at the Three Gorges by 5.0 m, decreased the excess flood volume by 56.2 billion m³, increased the peak flood discharges at Zhicheng and Chenglingji by 2 603 m³/s and 1 035 m³/s, respectively, and lowered the final water level of the Three Gorges Reservoir by 4.2 m. The coordinated scheduling scheme ensured that Qingjiang reservoir reached its highest water level before mid-July and maintained high-water-level operation until the flood season ended in the Yangtze River Basin, thereby fully achieving its flood peak shaving and detention functions. [Conclusion] This study proposes innovative strategies for activating flood control capacity in the Three Gorges and Qingjiang cascade reservoirs, specifically tailored to different regional flood compositions, thereby providing decision-makers with diversified scheduling schemes. The findings demonstrate that optimal scheduling schemes significantly enhance flood mitigation benefits, as evidenced by substantial reductions in excess flood volume and peak discharge. However, constrained by the limited flood control capacity of the reservoir group, the joint scheduling of Qingjiang and Three Gorges reservoirs cannot maximize flood control benefits across the river basin. Future efforts should incorporate additional water projects, such as flood detention areas, into the joint scheduling framework to further strengthen flood control capabilities across the river basin. This study provides theoretical foundations and technical support for exploring the coordinated scheduling potential of large-scale river basin reservoir groups and facilitating science-based decision-making.