水库生态调度优化模拟研究进展与展望

Research progress and prospects on the optimization simulation of reservoir ecological operation

摘要

为缓解水利工程运行对河流生态系统造成的负面影响，实施水库生态调度成为关键有效的手段。水库生态调度模型的类别主要分为生态约束型和生态目标型 2 类，其中，生态约束包括生态基流约束、关键物种需求约束、流域生态需求综合约束等，生态目标涵盖水文过程目标、栖息地目标、水温与水质目标等，并对比分析不同类型调度模型的机理、优劣点及应用场景。对比分析经典优化算法、进化算法、群体智能算法及混合算法的求解策略，阐述适用性与改进方向。针对当前研究存在的生态目标量化方法单一、多目标非劣解筛选主观性强等关键难题，提出基于负外部性评价与多维博弈的决策思路，以提升模拟的合理性，为水库生态调度研究从理论均衡迈向生态可持续的精准调控提供方向性参考。

Abstract

Reservoirs served as essential infrastructure for water resource utilization and regulation, providing significant benefits in flood control, hydropower generation, water supply, and navigation. However, the construction and operation of large-scale reservoirs significantly altered the natural hydrological regimes of rivers. These changes were observed as the flattening of flow processes, attenuation of flood peaks, stratification of water temperature, alteration of sediment transport, and variability in water mixing characteristics. These changes ultimately resulted in habitat degradation, obstruction of fish migration, biodiversity loss, and a decline in ecological functions. They also affected the geochemical behavior of nutrients and interfered with ecological processes. Against this background, the operation paradigm of reservoirs shifted from a focus on maximizing engineering benefits toward the emerging concept of ecological operation, which aimed to reconcile human demands with ecological requirements. A comprehensive review of recent research on simulating and optimizing reservoir ecology was conducted. The evolution of ecological-constraint-based and ecological-objective-based models was systematically traced. The former treated ecological flow as a constraint within traditional optimization frameworks, whereas the latter directly integrated ecological objectives into multi-objective models. A variety of solution strategies, including evolutionary algorithms, swarm intelligence algorithms, hybrid approaches, and classical optimization algorithms such as dynamic programming and its variations, were compared. Each algorithm's theoretical mechanism, computational efficiency, adaptability to complex systems, and limitations were carefully examined. Furthermore, attention was given to the generation and selection of Pareto-optimal solutions in multi-objective problems, with emphasis on the challenges of screening ecologically favorable solutions when confronted with economic objectives such as power generation and water supply. The review results indicated that ecological-constraint-based models maintained basic ecological flows while preserving economic benefits, but their contribution to ecosystem restoration was limited. Ecological-objective-based models improved responsiveness to ecological needs and allowed for a more holistic view of ecosystem processes, but quantifying ecological objectives presented significant methodological challenges. The "curse of dimensionality" in large-scale reservoir systems limited the theoretical rigor and global optimality that classical optimization techniques offered under simplified conditions. Strong global search capabilities and robustness in producing a variety of non-dominated solutions were demonstrated by evolutionary algorithms, especially genetic algorithms and their enhanced versions. Swarm intelligence algorithms, such as particle swarm optimization, achieved rapid convergence with fewer parameters but were prone to premature convergence. Hybrid algorithms, which combine mathematical programming, heuristic search, and local refinement, have shown promising results in large-scale and long-horizon scheduling problems. Overall, existing methods had achieved progress in balancing ecological and economic objectives, but challenges persisted in the accurate quantification of ecological demands, the scientific selection of non-dominated solutions, and the treatment of uncertainty from climate variability, hydrological forecasting errors, and socio-economic dynamics. Future research on reservoir ecological operation should address several unresolved issues. First, robust methods for the multidimensional quantification of ecological objectives are needed that go beyond simple indicators to capture complex ecological processes, such as spawning pulses, habitat dynamics, and water quality thresholds. Second, more systematic frameworks are necessary to screen Pareto-optimal solutions by evaluating negative externalities and ensuring that ecological objectives are not marginalized in decision-making. To address these shortcomings, a decision-making approach based on negative externality evaluation and multidimensional game theory was proposed, enabling the transition from a set of non-dominated solutions to the identification of an ecologically optimal solution. Finally, the application of large-scale models and digital twin technologies was envisaged to revolutionize ecological operation. While digital twins enable real-time interaction, scenario simulation, and adaptive optimization, large-scale models enable the coupling of physical mechanisms with data-driven insights. These technologies, combined with advances in remote sensing and sensor networks, are expected to create intelligent ecological operation systems capable of dynamic adjustment and risk prevention. Interdisciplinary collaboration among engineers, ecologists, policymakers, and stakeholders would be essential to ensure that future reservoir operation frameworks are technically sound, ecologically sustainable, and socially acceptable. Thus, this study provided theoretical guidance as well as practical opportunities for moving reservoir ecological operation toward a paradigm of efficiency, ecological integrity, and sustainability.

关键词

水库 / 生态调度 / 多目标优化 / 均衡调控 / 负外部性

Key words

reservoir / ecological operation / multi-objective optimization / balanced regulation / negative externality

参考文献

[1] 王何予, 田晶, 郭生练, 等. 考虑水文改变生态指标的丹江口水库多目标优化调度[J]. 南水北调与水利科技（中英文）, 2022, 20(6): 1041-1051. DOI: 10.13476/j.cnki.nsbdqk.2022.0103.

[2] 韩忠青, 刘招, 肖瑜, 等. 未来气候情景下引汉济渭工程水源区水文情势演变规律研究[J]. 水利水电技术（中英文）, 2025, 56(1): 26-38. DOI: 10.13928/j.cnki.wrahe.2025.01.003.

[3] 姜伟, 蒲艳, 邓华堂, 等. 溪洛渡-向家坝梯级水库生态调度对长江上游江津江段四大家鱼繁殖的影响[J]. 水生生物学报, 2025, 49(10): 38-47. DOI: 10.3724/1000-3207.2025.2024.0488.

[4] 陈金凤, 曾凌, 李雨, 等. 金沙江下游水温变化规律探讨[J]. 长江科学院院报, 2024, 41(6): 200-206.

[5] 陈庆伟, 刘兰芬, 刘昌明. 筑坝对河流生态系统的影响及水库生态调度研究[J]. 北京师范大学学报（自然科学版）, 2007(5): 578-582.

[6] 徐翔宇, 李心雨, 李慧, 等. 滦河流域水库群景观格局及生态风险时空变化[J]. 南水北调与水利科技（中英文）, 2025, 23(增刊 1): 159-164.

[7] 董哲仁. 筑坝河流的生态补偿[J]. 中国工程科学, 2006(1): 5-10.

[8] 卢有麟. 流域梯级大规模水电站群多目标优化调度与多属性决策研究[D]. 武汉: 华中科技大学, 2012.

[9] Castelletti A, Pianosi F, Soncini-Sessa R. Water reservoir control under economic, social and environmental constraints[J]. Automatica, 2008, 44(6): 1595-1607. DOI: 10.1016/j.automatica.2008.03.003.

[10] 张丹, 鲍军, 李想, 等. 考虑河流生态需水约束的调水规模研究[J]. 南水北调与水利科技（中英文）, 2023, 21(5): 917-929. DOI: 10.13476/j.cnki.nsbdqk.2023.0089.

[11] 艾学山, 范文涛. 水库生态调度模型及算法研究[J]. 长江流域资源与环境, 2008(3): 451-455. DOI: 10.3969/j.issn.1004-8227.2008.03.024.

[12] 梅亚东, 杨娜, 翟丽妮. 雅砻江下游梯级水库生态友好型优化调度[J]. 水科学进展, 2009, 20(5): 721-725. DOI: 10.14042/j.cnki.32.1309.2009.05.014.

[13] 尹正杰, 杨春花, 许继军. 考虑不同生态流量约束的梯级水库生态调度初步研究[J]. 水力发电学报, 2013, 32(3): 66-70.

[14] Richter B D, Baumgartner J V, Wigington R, et al. How much water does a river need?[J]. Freshwater Biology, 1997, 37： 231-249. DOI: 10.2307/jj.41003519.5.

[15] Sale M J, Brill E D, Herricks E E. An approach to optimizing reservoir operation for downstream aquatic resources[J]. Water Resources Research, 1982, 18(4): 705-712. DOI: 10.1029/WR018i004p00705.

[16] Richter B D, Baumgartner J V, Powell J, et al. A method for assessing hydrologic alteration within ecosystems[J]. Conservation Biology, 1996, 10： 1163-1174. DOI: 10.1046/j.1523-1739.1996.10041163.x.

[17] 赵越. 面向河流生境改善的水库调度建模理论与方法研究[D]. 武汉: 华中科技大学, 2014.

[18] 康玲, 黄云燕, 杨正祥, 等. 水库生态调度模型及其应用[J]. 水利学报, 2010, 41(2): 134-141. DOI: 10.13243/j.cnki.slxb.2010.02.008.

[19] 胡和平, 刘登峰, 田富强, 等. 基于生态流量过程线的水库生态调度方法研究[J]. 水科学进展, 2008(3): 325-332.

[20] Zhang Yuchen, Zhou Wenhua, Bai Tao, et al. Study on the impact of multi-source uncertainty on the operation of inter-basin water transfer projects and adaptive response strategies[J]. Journal of Hydrology, 2025, 660(PA): 133375. DOI: 10.1016/J.JHYDROL.2025.133375.

[21] 张睿, 李纪辉, 鲁春辉, 等. 考虑生态流量保证率的丹江口水库多目标优化调度研究[J]. 长江科学院院报, 2025, 42(9): 192-201.

[22] 谢雨祚, 尹家波, 李千珣, 等. 考虑生态流量的清江梯级水库优化调度研究[J]. 水资源研究, 2021, 10(1): 1-10. DOI: 10.12677/JWRR.2021.101001.

[23] Cardwell H, Jager H, Sale M. Designing instream flows to satisfy fish and human water needs[J]. Journal of Water Resources Planning and Management, 1996, 122(5): 356-363. DOI: 10.1061/(ASCE)0733-9496(1996)122:5(356).

[24] 周颖, 周研来, 鲁圆圆. 基于碳减排的梯级水库消落期多目标水碳调度研究[J]. 水生态学杂志, 2024, 45(1): 18-25. DOI: 10.15928/j.1674-3075.202310250297.

[25] 王奕博, 曾凌, 刘攀, 等. 面向水生群落系统稳定性的水库生态调度[J]. 水科学进展, 2024, 35(6): 914-926. DOI: 10.14042/j.cnki.32.1309.2024.06.005.

[26] 陈端, 陈求稳, 陈进. 考虑生态流量的水库优化调度模型研究进展[J]. 水力发电学报, 2011, 30(5): 248-256.

[27] Bellman R. Dynamic programming[J]. Science, 1966, 153: 34-37. DOI: 10.1126/science.153.3731.34.

[28] 王森, 程春田, 武新宇, 等. 梯级水电站群长期发电优化调度多核并行随机动态规划方法[J]. 中国科学: 技术科学, 2014, 44(2): 209-218.

[29] 郭生练, 陈炯宏, 刘攀, 等. 水库群联合优化调度研究进展与展望[J]. 水科学进展, 2010, 21(4): 496-503. DOI: 10.14042/j.cnki.32.1309.2010.04.013.

[30] 王少波, 解建仓, 孔珂. 自适应遗传算法在水库优化调度中的应用[J]. 水利学报, 2006(4): 480-485. DOI: 10.3969/j.issn.1008-1402.2026.02.022.

[31] 刘攀, 郭生练, 李玮, 等. 遗传算法在水库调度中的应用综述[J]. 水利水电科技进展, 2006(4): 78-83. DOI: 10.3880/j.issn.1006-7647.2006.04.024.

[32] 雷其鸣, 牛庚, 桑学锋, 等. 基于改进强化学习算法的水库群多目标优化调度[J]. 南水北调与水利科技（中英文）, 2025, 23(5): 1185-1195. DOI: 10.13476/j.cnki.nsbdqk.2025.0118.

[33] 刘东, 黄强, 杨元园, 等. 基于改进 NSGA-Ⅱ算法的水库双目标优化调度[J]. 西安理工大学学报, 2020, 36(2): 176-181. DOI: 10.19322/j.cnki.issn.1006-4710.2020.02.006.

[34] 谢云东, 章四龙, 王红瑞, 等. 基于 NSGA-Ⅱ算法的硗碛水电站多目标调度研究[J]. 中国农村水利水电, 2022(3): 207-211.

[35] 白涛, 阚艳彬, 畅建霞, 等. 水库群水沙调控的单-多目标调度模型及其应用[J]. 水科学进展, 2016, 27(1): 116-127. DOI: 10.14042/j.cnki.32.1309.2016.01.013.

[36] 黄显峰, 王宁, 刘志佳, 等. 基于改进 NSGA-Ⅱ算法的梯级水库多目标优化调度[J]. 水利水电科技进展, 2024, 44(4): 51-58. DOI: 10.3880/j.issn.1006-7647.2024.04.008.

[37] Jain H, Deb K. An evolutionary many-objective optimization algorithm using reference-point based non-dominated sorting approach, part Ⅰ: Solving problems with box constraints[J]. IEEE Transactions on Evolutionary Computation, 2014, 18(4): 577-601. DOI: 10.1109/TEVC.2013.2281535.

[38] Jain H, Deb K. An evolutionary many-objective optimization algorithm using reference-point based non-dominated sorting approach, part Ⅱ: Handling constraints and extending to an adaptive approach[J]. IEEE Transactions on Evolutionary Computation, 2014, 18(4): 602-622. DOI: 10.1109/TEVC.2013.2281534.

[39] 唐晓宇, 刘铁, 黄粤, 等. 开都河流域水-能源-生态综合收益下的水资源优化配置[J]. 南水北调与水利科技（中英文）, 2025, 23(1): 90-98. DOI: 10.13476/j.cnki.nsbdqk.2025.0009.

[40] 何英, 唐晓宇, 彭亮, 等. 南疆丰收灌区水资源多目标优化配置方案优选[J]. 农业工程学报, 2021, 37(6): 117-126. DOI: 10.11975/j.issn.1002-6819.2021.06.015.

[41] 唐晓宇. 新疆阿瓦提灌区水资源多目标优化及方案优选[D]. 乌鲁木齐: 新疆农业大学, 2021. DOI: 10.27431/d.cnki.gxnyu.2021.000026.

[42] 李弘瑞. 基于多目标进化算法的黄河中游水库群优化调度[D]. 郑州: 华北水利水电大学, 2022. DOI: 10.27144/d.cnki.ghbsc.2022.000147.

[43] 李传利. 基于 NSGA-Ⅲ的黄河上游梯级水库水沙联合优化调度研究[D]. 郑州: 华北水利水电大学, 2024. DOI: 10.27144/d.cnki.ghbsc.2024.000102.

[44] 丁航, 安源, 王颂凯, 等. 水光互补的短期优化调度[C]// 2016 第二届能源, 环境与地球科学国际会议论文集. 陕西, 2016: 21-26.

[45] 李豹, 周建中, 欧阳硕. 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[J]. 水资源研究, 2012, 1(3): 45-51. DOI: 10.12677/JWRR.2012.13007.

[46] 杨光, 郭生练, 刘攀, 等. PA-DDS 算法在水库多目标优化调度中的应用[J]. 水利学报, 2016, 47(6): 789-797. DOI: 10.13243/j.cnki.slxb.20150773.

[47] Tang Zhengyang, Liu Shuai, Qin Hui, et al. Chaos-enhanced Harris Hawks optimizer for cascade reservoir operation with ecological flow similarity[J]. Sustainability, 2025, 17(19): 8616. DOI: 10.3390/SU17198616.

[48] Feng Zhongkai, Niu Wenjing, Cheng Chuntian. Optimization of hydropower reservoirs operation balancing generation benefit and ecological requirement with parallel multi-objective genetic algorithm[J]. Energy, 2018, 153： 706-718. DOI: 10.1016/j.energy.2018.04.075.

[49] 黄显峰, 吴志远, 李昌平, 等. 基于改进粒子群-逐次逼近法的水库调度图多目标优化[J]. 水利水电科技进展, 2021, 41(2): 1-7.

[50] Mao Jingqiao, Tian Mingming, Hu Tengfei, et al. Shuffled complex evolution coupled with stochastic ranking for reservoir scheduling problems[J]. Water Science and Engineering, 2019, 12(4): 307-318. DOI: 10.1016/j.wse.2019.12.003.

[51] 强安丰, 汪妮, 莫淑红, 等. 气候变化对水文水资源影响评价的不确定研究进展[J]. 水资源研究, 2020, 9(2): 169-178. DOI: 10.12677/JWRR.2020.92018.

[52] 桑学锋, 刘志武, 王浩, 等. 长江水资源调配推演器构建与推演分析[J]. 中国水利, 2024(1): 6-11. DOI: 10.3969/j.issn.1000-1123.2024.01.004.

[53] 陈晓楠, 顾起豪, 靳燕国, 等. 南水北调中线输水调度预演系统研发及应用[J]. 中国水利, 2024(1): 12-17. DOI: 10.3969/j.issn.1000-1123.2024.01.005.

基金资助

国家自然科学基金项目(U2240202)

国家自然科学基金项目(52394233)

中国水利水电科学研究院基本科研业务费资助项目(WR110145B0032025)

流域水循环与水安全全国重点实验室自主研究项目(WR110146B0022024)

流域水循环与水安全全国重点实验室自主研究项目(SKL2025TDGG06)

流域水循环与水安全全国重点实验室自主研究项目(SKL2025RCPY09)