基于决策因子选择的梯级水库多目标优化调度规则研究

发布时间：2018-05-11 08:02

本文选题：梯级水库调度 + 多目标优化　；参考：《水利学报》2017年08期

【摘要】：运用极端随机森林挖掘水库调度信息,考虑各水库相关因子与决策目标的相互关系,提出了梯级水库输入数据选择(Cascade-reservoir Input-variables Selection,CIS)方法,引入Gaussian径向基函数建立水库调度规则,并采用PA-DDS多目标算法对水库调度规则参数进行优化,得到了同时考虑供水和发电的多目标优化调度规则集。汉江上游梯级水库的应用结果表明:CIS方法能优选得到较为稳定的决策因子,且能在充分利用已有优选信息的同时避免决策因子选取上的重复和冗余;经梯级多目标优化调度,优选的决策因子能得到比传统决策因子结果分布更优的非劣解集,在以发电为主要目标进行优化调度时,能有效调整汛期和非汛期的水量利用方式,梯级水库年均发电量和供水量分别提高了0.33%和7.84%。
[Abstract]:By using extremely random forest mining reservoir operation information and considering the relationship between the relative factors of each reservoir and the decision object, the Cascade-reservoir Input-variables selection method is proposed, and the Gaussian radial basis function is introduced to establish the reservoir operation rules. PA-DDS multi-objective algorithm is used to optimize the parameters of reservoir dispatching rules, and a set of multi-objective optimal dispatching rules considering both water supply and power generation is obtained. The application results of cascade reservoirs in the upper reaches of the Hanjiang River show that the proportion CIS method can obtain more stable decision factors, and avoid duplication and redundancy in the selection of decision factors while making full use of the available optimal selection information. The optimal decision factors can get a better distribution of non-inferior solution set than the traditional decision factors. When the power generation is the main objective of optimal scheduling, the water use mode of flood season and non-flood season can be adjusted effectively. The average annual output and water supply of cascade reservoirs increased by 0.33% and 7.84% respectively.
【作者单位】：武汉大学水资源与水电工程科学国家重点实验室水资源安全保障湖北省协同创新中心;
【基金】：国家自然科学基金重点项目(51539009) 十三五国家重点研发项目(2016YFC0402206)
【分类号】：TV697.1

【相似文献】