摘要
预报年最大洪峰流量对水库安全运行、科学调度具有重要的指导意义。研究以平寨水库为研究对象,基于国家气候中心提供的130项气象因子,采用逐步多元回归、神经网络、支持向量机、决策树4种方法,建立了平寨水库的年最大洪峰流量预报模型。预报精度的分析比较结果表明,4种模型对平寨水库年最大洪峰流量预报都表现出较好的模拟预报效果,其中以神经网络和逐步多元回归这两种模型精度最高,在制订平寨水库汛期安全度汛计划时具有一定参考价值;也表明了基于气象因子预报年最大洪峰流量是可行的。
The forecasting of annual maximum flood peak flow is of great significance to the safe and scientific operation of reservoirs.Taking Pingzhai Reservoir as the research object,the annual maximum flood peak flow model is studied and established by using four methods including stepwise multiple regression,neural network,support vector machine and decision tree based on the data of 130 meteorological factors provided by the National Climate Center.The prediction accuracy of the four methods is analyzed and compared,and the result shows that the four methods all have good simulation and prediction effects on the annual maximum flood peak flow of Pingzhai Reservoir,and among them,the neural network and stepwise multiple regression models have the highest accuracy,which has certain reference value when formulating the flood control plan of Pingzhai Reservoir.The result also shows that it is feasible to forecast the annual maximum flood peak flow based on meteorological factors.
作者
石庆安
李意
罗天文
王茂洋
赵朝彬
张健源
SHI Qing an;LI Yi;LUO Tianwen;WANG Maoyang;ZHAO Chaobin;ZHANG Jianyuan(Qiandongnan Water Conservancy Investment(Group)Co.,Ltd.,Kaili 522601,Guizhou,China;Guizhou Water Conservancy and Hydropower Survey Design and Research Institute Co.,Ltd.,Guiyang 550002,Guizhou,China)
出处
《水力发电》
CAS
2021年第11期1-3,25,共4页
Water Power
基金
贵州省重大科技专项(黔科合重大专项字[2017]3005号)
贵州省水利科技项目(KT201903)。
关键词
年最大洪峰
预报模型
逐步多元回归
神经网络
支持向量机
决策树
平寨水库
annual maximum flood peak
forecasting model
stepwise multiple regression
neural network
support vector machine
decision tree
Pingzhai Reservoir
