Reservoir flood control operation is an important and effective measure for avoiding or reducing flood disasters. Considering that reservoir operation is a control process that essentially manages the spillway gates to increase or decrease the amount of released water, this paper focuses on studying the reservoir flood control operation which considers spillway gate scheduling. Minimizing the peak outflow from the reservoir is adopted as the optimization objective to protect downstream areas from flood disasters. In addition to the conventional reservoir constraints, the operation constraints of individual spillway gates and the influence of the reservoir forebay water level and opening degree on the gates’ discharge capacity are also well considered to obtain precise scheduling. The nonlinear factors, especially the objective function in the form of maximin, the three-dimensional water discharge curves, and the rules for the opening and closing sequence of the spillway gates, are linearized through several strategies. The original nonlinear and non-convex model is then converted into a standard mixed-integer linear programming (MILP) formulation. The results for the real-world case study indicate that: 1) the proposed model is believed to be computationally efficient with a calculation time of about 30 min; 2) for the flood events with return periods of 10 years and 100 years, the flood peaks are reduced from 4160 m³/s to 3681 m³/s and from 11007 m³/s to 9293 m³/s, representing decreases of 11.5 % and 15.6 %, respectively. The outflow curves become more stable, demonstrating that the flood control pressure on the downstream had been relieved after optimization; 3) the proposed model can produce more realistic and executable flood control operation scheduling compared with the conventional optimization model without considering spillway gate scheduling. These research findings can provide an effective reference for both the operators of reservoirs and researchers in the field of reservoir flood control operation.

水库防洪作业是避免或减轻洪水灾害的一项重要而有效的措施。考虑到水库运行本质上是一个控制溢洪道闸门以增加或减少泄水量的控制过程，本文重点研究考虑溢洪道闸门调度的水库防洪运行。将水库的峰值出流最小化作为保护下游地区免受洪水灾害的优化目标。除了常规的水库约束外，还充分考虑了各个溢洪道闸门的运行约束以及水库前湾水位和开度对闸门泄洪能力的影响，以获得精确调度。非线性因素，特别是maxmin形式的目标函数，三维排水曲线，以及溢洪道闸门的开启和关闭顺序规则，通过多种策略线性化。然后将原始非线性和非凸模型转换为标准混合整数线性规划 (MILP) 公式。实际案例研究的结果表明：1) 所提出的模型被认为具有计算效率，计算时间约为 30 分钟；2) 对于重现期为 10 年和 100 年的洪水事件，洪峰从 4160 m 降低 然后将原始非线性和非凸模型转换为标准混合整数线性规划 (MILP) 公式。实际案例研究的结果表明：1) 所提出的模型被认为具有计算效率，计算时间约为 30 分钟；2) 对于重现期为 10 年和 100 年的洪水事件，洪峰从 4160 m 降低 然后将原始非线性和非凸模型转换为标准混合整数线性规划 (MILP) 公式。实际案例研究的结果表明：1) 所提出的模型被认为具有计算效率，计算时间约为 30 分钟；2) 对于重现期为 10 年和 100 年的洪水事件，洪峰从 4160 m 降低³ /s 到 3681 m ³ /s 和从 11007 m ³ /s 到 9293 m ³ /s，分别下降 11.5 % 和 15.6 %。出流曲线趋于平稳，说明优化后下游防洪压力得到缓解；3) 与常规优化模型相比，该模型在不考虑泄洪闸调度的情况下，可以产生更加现实和可执行的防洪调度调度。这些研究成果可为水库运营者和水库防洪作业领域的研究人员提供有效的参考。