多目标粒子群算法在梯级水库联合防洪调度中的应用研究

Application of Multi-Objective Particle Swarm Algorithm in Joint Flood Control Operation of Cascade Reservoirs

摘要: 针对金沙江下游梯级与三峡梯级联合防洪调度中的多目标优化问题，综合考虑各水库坝身防洪安全及上下游防护区安全等目标，建立了梯级水库联合防洪多目标优化调度模型，提出多目标粒子群算法对模型进行优化求解，构造外部档案集来保存求解过程中的非劣解，并基于个体聚集距离来维持外部档案集的粒子分布性，得到了分布均匀且范围广的非劣调度方案集供决策者评价优选。分别选取1981年和1998年为典型洪水年，对各典型年不同频率下的设计洪水进行了联合防洪优化调度，实例研究结果表明，相比于三峡梯级单独运行，梯级水库联合防洪调度可有效削减三峡梯级的入库洪峰流量，降低三峡坝前最高水位，提高长江中下游的防洪能力。

Abstract: Focusing on the multi-objective problem of joint flood control operation for the JinshaRiverdownstream cascade reservoirs and the Three Gorges cascade reservoirs, a multi-objective optimization model of the joint flood control operation for Cascade Reservoirs was established, considering the objectives of flood control safety of cascade reservoirs and upstream and downstream protection zones. The model was solved by the proposed multi-objective particle swarm algorithm. Meanwhile, the set of external files to save the non-inferior solutions in the solution process was constructed, based on the clustering distance of the particle which maintains the distribution of the external file set. Finally, the solution distributes of a wide range set of the non-inferior scheduling programs offering to policy makers for evaluation and selection. 1981 and 1998 were selected as the typical years, a joint control optimization scheduling under design flood has been done at different frequencies in the typical flood year. And cases study results has shown that, compared with a separate run in the Three Gorges cascade, the joint flood control of cascade reservoirs can effectively reduce the Three Gorges Cascade storage peak flow and the highest water level of the dam in Three Gorges, meanwhile, enhance the flood control capacity of the middle and lower reaches of the Yangtze River.