杨哲¹, 杨侃¹, 夏怡¹, 马一鸣¹, 黄悦¹, 许军良²
AuthorsHTML:杨哲¹, 杨侃¹, 夏怡¹, 马一鸣¹, 黄悦¹, 许军良²
AuthorsListE:Yang Zhe¹, Yang Kan¹, Xia Yi¹, Ma Yiming¹, Huang Yue¹, Xu Junliang²
AuthorsHTMLE:Yang Zhe¹, Yang Kan¹, Xia Yi¹, Ma Yiming¹, Huang Yue¹, Xu Junliang²
Unit:1. 河海大学水文水资源学院，南京 210098；2. 宜兴市周铁镇水利农机站，宜兴 214200
Unit_EngLish:1. College of Hydrology and Water Resources, Hohai University, Nanjing 210098, China
2. Water Conservancy and Agricultural Municipal Station of Zhoutie, Yixing 214200, China
Abstract_Chinese:为协调清江梯级水库发电及下游生态效益, 实现梯级库群水能资源高效利用和生态效益双赢, 本文对清江隔河沿、高坝洲控制断面基础、适宜和理想等级生态流量进行分析计算, 并以发电量和生态流量贴近度最优作为目标函数, 建立兼顾生态保护的清江梯级库群多目标优化调度模型．进一步, 针对传统混合蛙跳算法(SFLA)存在搜索能力有限、易陷入局部最优等问题, 从混沌种群初始化、局部搜索方式、基于云模型理论精英青蛙进化和后期启发式激活机制等方面对SFLA进行改进, 提出基于混沌云模型理论混合蛙跳算法(CNSFLA)．采用该方法求解梯级水库生态调度模型, 长系列优化调度结果显示, 在保证发电效益和下游控制断面基础生态流量的同时, 有效提升其适宜、理想生态流量保证率, 总体保证率维持在较高水平; 对于枯水典型年, 虽然通过算法优化调度, 但天然入流偏少导致控制断面部分时段理想生态流量仍无法满足．长系列调度结果对比表明CNSFLA具备高效搜索性能和良好稳定性, 可显著提高库群联合调度结果质量．
Abstract_English:In order to coordinate power generation and downstream ecological benefits，and achieve the win-win goal of efficient water resources exploitation in Qingjiang cascade reservoirs，the basic，proper and ideal ecological flow of Geheyan and Gaobazhou control section is analyzed and calculated in this paper. Moreover，the power generation and ecological flow simulation degree is set as the objective function to establish the multi-objective scheduling model in consideration of ecological protection in Qingjiang cascade reservoirs. In addition，as to the problems in traditional shuffled frog leaping algorithm(SFLA)such as poor search capability and easily falling into local optimum，an improved CNSFLA based on chaos and cloud model is proposed to overcome these defects. The improvements can be embodied as chaos population initialization，updated local search method，evolution strategy for elite frog and heuristic activation mechanism during later iteration. Then，the improved method is applied to ecological scheduling model of cascade reservoirs. The scheduling results for long series indicates that the benefits of electricity and basic ecological flow in downstream control section are guaranteed with efficient promotion and high level of suitable and ideal ecological flow guarantee. In terms of a typical dry year，the less inflow will make it hard to reach the ideal ecological flow demand for the control section even if the scheduling is optimized by CNSFLA. Finally，comparison results for long series demonstrate the effective search ability with good stability，and high quality of scheduling results.
Keyword_Chinese:混合蛙跳算法(SFLA); 梯级库群多目标优化调度; 混沌种群初始; 云模型理论; 精英进化; 启发式青蛙激活机制
Keywords_English:shuffled frog leaping algorithm(SFLA); multi-objective optimal scheduling of cascade reservoirs; chaos population initialization; cloud model; elite evolution; heuristic frog activation mechanism

PDF全文下载地址:http://xbzrb.tju.edu.cn/#/digest?ArticleID=6145