基于混合智能算法的梯级水库群日优化调度研究  

Research on Daily Optimal Dispatch of Cascade Reservoirs Based on Hybrid Intelligent Algorithm

作　　者：刘喜峰 LIU Xi-feng(School of Mechanical and Civil Engineering,Jilin Agricultural Science and Technology University,Jilin 132101,China)

机构地区：[1]吉林农业科技学院机械与土木工程学院,吉林吉林132101

出　　处：《水电能源科学》2021年第10期76-80,共5页Water Resources and Power

基　　金：吉林省教育厅科学技术研究项目(JJKH20210417KJ)。

摘　　要：针对梯级水库间水流滞时导致的水库群短期优化调度水量不平衡的问题,提出一种动态水流滞时的梯级水电系统日优化调度模型。先通过人工神经网络智能算法训练水库群关键数据,实现对动态水流滞时函数的动态拟合;其次构建梯级水电站水库群日优化调度模型,通过逐次优化算法对调度模型进行求解;最后选取锦东、官地水电站对固定水流滞时和动态水流滞时开展试验验证与分析。结果表明,相较于固定水流滞时,动态水流滞时能更加精准地描述梯级水库群间的水流关系,且能有效提高梯级水电站水库群的发电效率和经济效益。Aiming at the imbalance of the short-term optimal dispatch of reservoir groups caused by the delay time of water flow among cascade reservoirs,a daily optimal dispatch model of cascade hydropower system with dynamic flow delay time is proposed.Firstly,the artificial neural network intelligent algorithm was used to train the key data of the reservoir group to realize the dynamic fitting of the dynamic flow lag time function.Secondly,the daily optimal dispatch model of cascade hydropower stations was established,and the dispatch model was solved by progressive optimality algorithm.Finally,Jindong and Guandi Hydropower Stations were selected to carry out experimental verification and analysis on fixed and dynamic water delay time.The experimental results show that the dynamic water delay time can describe the flow relationship between cascade reservoirs more accurately than the fixed water delay time,and can effectively improve the power generation efficiency and economic benefits of cascade hydropower station reservoirs.

关 键 词：水流滞时 梯级水库 优化调度 混合智能算法 水电站 

分 类 号：TV697[水利工程—水利水电工程]