Integration of deep learning and improved multi-objective algorithm to optimize cascade reservoirs operation with consideration of ecological dissolved oxygen needs
文献类型：期刊论文
| 作者 | Zhu, Zhaoyang1; Li, Haoran1; Wang, Zhaocai1; Zhang, Xingxing2; Tan, Zuowen1 |
| 刊名 | JOURNAL OF HYDROLOGY
 |
| 出版日期 | 2026-03-01 |
| 卷号 | 667页码:134899 |
| 关键词 | Cascade reservoirs Multi-objective trade-off Ecological dissolved oxygen (DO) Deep learning Pareto front |
| ISSN号 | 0022-1694 |
| DOI | 10.1016/j.jhydrol.2025.134899 |
| 产权排序 | 2 |
| 文献子类 | Article |
| 英文摘要 | In the context of global warming and the development of hydraulic projects, changes in hydrology may affect downstream water quality, thereby increasing the risk of algal blooms. In reservoir ecological operation management, there has been a long-standing issue of neglecting the impact of flow variations on downstream dissolved oxygen (DO) levels. Although previous studies have incorporated ecological flow requirements into reservoir operation strategies to balance power generation, flood control, and ecological protection needs, comprehensive trade-off analyses remain insufficient. To bridge this gap, the study proposes a coupled framework (DL-CRS) that integrates a deep learning (DL) model with a cascade reservoir scheduling (CRS) model to optimize complex scheduling problems. The study is applied to a group of lower Jinsha River (JSR) cascade reservoirs, focusing on DO needs at the Hengjiang River (HJR) inlet. This research constructs a CNN-BiLSTM model to predict DO changes accurately and proposes a multi-strategy enhanced algorithm to improve the quality of the Pareto solution set. The results demonstrate that the optimized scheme improves upon the conventional approach by 6.40 % in power generation, 12.93 % in flood control, and 7.27 % in ecological benefits across various typical year scenarios. This study can provide decision support for intelligent scheduling and water quality safety assurance of cascade reservoirs group. |
| URL标识 | 查看原文 |
| WOS研究方向 | Engineering ; Geology ; Water Resources |
| 语种 | 英语 |
| WOS记录号 | WOS:001664717100001 |
| 出版者 | ELSEVIER |
| 源URL | [http://ir.igsnrr.ac.cn/handle/311030/219611]  |
| 专题 | 资源与环境信息系统国家重点实验室_外文论文 |
| 通讯作者 | Wang, Zhaocai |
| 作者单位 | 1.Shanghai Ocean Univ, Coll Informat, Shanghai 201306, Peoples R China; 2.Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, Beijing 100101, Peoples R China |
| 推荐引用方式 GB/T 7714 | Zhu, Zhaoyang,Li, Haoran,Wang, Zhaocai,et al. Integration of deep learning and improved multi-objective algorithm to optimize cascade reservoirs operation with consideration of ecological dissolved oxygen needs[J]. JOURNAL OF HYDROLOGY,2026,667:134899. |
| APA | Zhu, Zhaoyang,Li, Haoran,Wang, Zhaocai,Zhang, Xingxing,&Tan, Zuowen.(2026).Integration of deep learning and improved multi-objective algorithm to optimize cascade reservoirs operation with consideration of ecological dissolved oxygen needs.JOURNAL OF HYDROLOGY,667,134899. |
| MLA | Zhu, Zhaoyang,et al."Integration of deep learning and improved multi-objective algorithm to optimize cascade reservoirs operation with consideration of ecological dissolved oxygen needs".JOURNAL OF HYDROLOGY 667(2026):134899. |
入库方式： OAI收割
来源：地理科学与资源研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。